CN108734186A

CN108734186A - Automatically exit from the methods, devices and systems of instant communication session group

Info

Publication number: CN108734186A
Application number: CN201710254343.4A
Authority: CN
Inventors: 靳玉康
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2017-04-18
Filing date: 2017-04-18
Publication date: 2018-11-02
Anticipated expiration: 2037-04-18
Also published as: CN108734186B

Abstract

This application discloses a kind of methods, devices and systems automatically exiting from instant communication session group.Wherein, this method includes：At least one session group is traversed, the user behavior characteristics for each user for including in session group are obtained；According to the user behavior characteristics for each user for including in session group, the probability that each user exits session group is obtained；According to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.Present application addresses existing instant messaging tools the technical issues of moving back group operation cumbersome influence user experience effect.

Description

Automatically exit from the methods, devices and systems of instant communication session group

Technical field

This application involves field of communication technology, in particular to a kind of method automatically exiting from instant communication session group, Device and system.

Background technology

With the fast development of Internet technology, instant messaging has developed into collection exchange, information, amusement, search, electronics The synthesization information platform that commercial affairs, office cooperation and corporate client service etc. are integrated.User can be by being mounted with Instant Messenger Believe that the intelligent communications equipment such as computer or the mobile phone of software (for example, QQ, wechat, MSN etc.) and other have been mounted with respective client Hold the computer or mobile phone messaging of software.During the use of immediate communication tool (for example, wechat), especially enterprise is When communication tool in, often established according to need of work and many interim link up groups and more and more face with the progress of work When link up group can bring the problems such as issue in focus, simple interface.

Existing instant messaging tools mainly supports that user exits group by hand or timing is dismissed group's two ways and faced to exit When link up group.The former influences user's operation experience due to needing user to exit group chat by hand, in addition, when user exits group by hand, The prompt for oneself moving back group can be shown in group, pickle is brought to user itself；The latter due to being provided with group life cycle, Group can dismiss automatically within a certain period of time, if continuing communication if necessary, it is necessary to extend group's life cycle or build again Group, this is larger to the whole customer impact of group.

For above-mentioned problem, currently no effective solution has been proposed.

Invention content

An embodiment of the present invention provides a kind of methods, devices and systems automatically exiting from instant communication session group, at least Solve the technical issues of moving back group operation cumbersome influence user experience effect of existing instant messaging tools.

One side according to the ... of the embodiment of the present invention provides a kind of method automatically exiting from instant communication session group, packet It includes：At least one session group is traversed, the user behavior characteristics for each user for including in session group are obtained；It is wrapped according in session group The user behavior characteristics of each user contained obtain the probability that each user exits session group；According to the probability for exiting session group, Determine whether that the prompt message of session group is exited in output.

Another aspect according to the ... of the embodiment of the present invention additionally provides a kind of system automatically exiting from instant communication session group, Including：Server, for providing model file, wherein model file has pre-saved the session of at least one client feedback The user behavior characteristics of group, the user behavior characteristics of session group include the user behavior characteristics of at least one user in session group； Client obtains the user's row for each user for including in session group with server communication for traversing at least one session group It is characterized, and according to the user behavior characteristics for each user for including in session group, obtains each user and exit the general of session group Rate, according to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.

Another aspect according to the ... of the embodiment of the present invention additionally provides a kind of device automatically exiting from instant communication session group, Including：First acquisition module obtains the user behavior for each user for including in session group for traversing at least one session group Feature；Second acquisition module obtains each user and moves back for the user behavior characteristics according to each user for including in session group Go out the probability of session group；Determining module, for according to the probability for exiting session group, it is determined whether the prompt of session group is exited in output Information.

Another aspect according to the ... of the embodiment of the present invention additionally provides a kind of method exiting group, including：It obtains in first group Including multiple users user behavior characteristics；According to the user behavior characteristics of multiple users, obtains at least one user and exit Group of probability；According to probability, it is determined whether group of prompt message is exited in output.

Another aspect according to the ... of the embodiment of the present invention additionally provides a kind of device exiting group, including：First obtains list Member, the user behavior characteristics for obtaining in first group the multiple users for including；Second acquisition unit, for according to multiple users User behavior characteristics, obtain at least one user and exit group of probability；Determination unit, for according to probability, determination to be Group of prompt message is exited in no output.

Another aspect according to the ... of the embodiment of the present invention additionally provides storage medium, and storage medium includes the program of storage, In, what equipment where controlling storage medium when program is run executed above-mentioned any one automatically exits from instant communication session group's Method.

Another aspect according to the ... of the embodiment of the present invention additionally provides a kind of processor, which is characterized in that processor is for transporting Line program, wherein program executes the method for automatically exiting from instant communication session group of above-mentioned any one when running.

Another aspect according to the ... of the embodiment of the present invention, additionally provides system, which is characterized in that including：Processor；And it deposits Reservoir is connect with processor, for providing the instruction for handling following processing step for processor：Step 402, it traverses at least one Session group obtains the user behavior characteristics for each user for including in session group；Step 404, every according to include in session group The user behavior characteristics of a user obtain the probability that each user exits session group；Step 406, according to exiting the general of session group Rate, it is determined whether the prompt message of session group is exited in output.

In embodiments of the present invention, by traversing at least one session group, each user for including in session group is obtained User behavior characteristics；According to the user behavior characteristics for each user for including in session group, obtains each user and exit session group Probability；According to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output, has reached and has been existed according to user User behavior characteristics in session group predict that user exits the probability of the session group, to determine whether to prompt user to execute move back group The purpose of operation moves back group operation to realize simplified user's execution, improves the technique effect of user experience, and then solve existing There is the technical issues of moving back group operation cumbersome influence user experience effect of instant messaging tools.

Description of the drawings

Attached drawing described herein is used for providing further understanding of the present application, constitutes part of this application, this Shen Illustrative embodiments and their description please do not constitute the improper restriction to the application for explaining the application.In the accompanying drawings：

Fig. 1 (a) is " wechat " group chat list schematic diagram according to a kind of optional mobile phone version of the embodiment of the present application；

Fig. 1 (b) is to enter some group according to a kind of " wechat " group chat of optional mobile phone version of the embodiment of the present application Schematic diagram；

Fig. 1 (c) is to exit some group according to a kind of " wechat " group chat of optional mobile phone version of the embodiment of the present application Operation chart；

Fig. 2 is a kind of optional instant messaging network architecture schematic diagram according to the embodiment of the present application；

Fig. 3 is a kind of system schematic automatically exiting from instant communication session group according to the embodiment of the present application；

Fig. 4 is the flow chart according to a kind of method automatically exiting from instant communication session group of the embodiment of the present application；

Fig. 5 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Fig. 6 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Fig. 7 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Fig. 8 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Fig. 9 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 10 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 11 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 12 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 13 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 14 is to map schematic diagram according to a kind of optional data of the embodiment of the present application；

Figure 15 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 16 is the flow according to a kind of optional method for automatically exiting from instant communication session group of the embodiment of the present application Figure；

Figure 17 is a kind of schematic device automatically exiting from instant communication session group according to the embodiment of the present application；

Figure 18 is a kind of method flow diagram exiting group according to the embodiment of the present application；

Figure 19 is a kind of schematic device exiting group according to the embodiment of the present application；And

Figure 20 is the hardware block diagram according to a kind of terminal of the embodiment of the present application.

Specific implementation mode

In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application Attached drawing, technical solutions in the embodiments of the present application are clearly and completely described, it is clear that described embodiment is only The embodiment of the application part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people The every other embodiment that member is obtained without making creative work should all belong to the model of the application protection It encloses.

It should be noted that term " first " in the description and claims of this application and above-mentioned attached drawing, " Two " etc. be for distinguishing similar object, without being used to describe specific sequence or precedence.It should be appreciated that using in this way Data can be interchanged in the appropriate case, so as to embodiments herein described herein can in addition to illustrating herein or Sequence other than those of description is implemented.In addition, term " comprising " and " having " and their any deformation, it is intended that cover It includes to be not necessarily limited to for example, containing the process of series of steps or unit, method, system, product or equipment to cover non-exclusive Those of clearly list step or unit, but may include not listing clearly or for these processes, method, product Or the other steps or unit that equipment is intrinsic.

First, the part noun or term occurred during the embodiment of the present application is described is suitable for following solution It releases：

Bayes' theorem, also referred to as Bayesian inference, early in 18th century, British scholar Bayes (1702~1763) it is proposed that The formula of design conditions probability is used for solving the problems, such as following one kind：Assuming that H [1], H [2] ..., H [n] mutual exclusions and composition one are completely Event, it is known that their probability P (H [i]), i=1,2 ..., n now observe that certain event A and H [, 1], H [, 2] ..., and H [, n] and phase Occur with machine, and known conditions probability P (A/H [, i]), seeks P (H [, i]/A).

Euclidean distance, also known as euclidean metric (euclidean metric) are the distance definitions of a generally use, Refer to the actual distance in m-dimensional space between two points, or the natural length (i.e. the distance of the point to origin) of vector.Two Euclidean distance in peacekeeping three dimensions is exactly the actual range between 2 points.

It can be that is created in instant messaging tools include the group of multiple users that session group, which can be session group, can be with More human hair sending voices, video, picture or word are supported come the chat that conversates, for example, QQ groups, QQ discussion group, wechat group etc..

Group's state feature can be the feature in instant communication software for characterizing each session group state.

User behavior characteristics can basis for characterizing each behavioural characteristic of the user in the session group in session group Group's state feature of session group obtain each user user behavior characteristics.

Embodiment 1

According to the embodiment of the present application, a kind of system embodiment automatically exiting from instant communication session group, this implementation are provided Example can be applied to the management and group in the various immediate communication tools based on internet communication, including but not limited to QQ Group, wechat group, MSN groups, the application scenarios such as QQ discussion group.

Instant messaging (Instant messaging, abbreviation IM) is a terminal service Internet-based, allows two people Or more people are exchanged using the instant transmission text information of network, voice with video.Currently, multi-party meeting is all supported in instant messaging service Words, i.e. group chat.Inventor has found, since anyone could set up group, in particular, the QQ discussion group occurred in recent years, wechat Interim Qun Deng interim conversations group, creates group chat more operated quickly and conveniently, often in order to everybody is allowed to exchange some opinions or communication Some items (for example, time and location etc. of party are discussed) and establish a group temporarily so that the session group that user possesses is got over Come more.Since, many interim groups may use primary do not need again then or some interim groups are to be forced to draw in , by taking " wechat " of mobile phone version as an example, Fig. 1 (a) is " wechat " group chat list according to a kind of mobile phone version of the embodiment of the present application Schematic diagram contains that " product department ", " party discuss ", " corporate HQ ", " senior middle school is same as shown in Fig. 1 (a) in the group chat interface Learn ", multiple groups such as " red packet group " and " mobile phone buy exchange ".Group's " party discusses " may be it can be seen from Fig. 1 (a) The interim discussion group that user establishes or is added for certain this party, group's " mobile phone purchase exchange " may be user in purchase hand The group that the machine stage is added, for these interim groups for establishing or being added temporarily, group chat content may be with the relationship of user not Greatly, at this time, succinct for interface of keeping in communication, usual user can select to exit some useless groups.

Currently, the session group in existing instant messaging tools, still needs to execute by hand and moves back group operation, and different operation systems It is also different to move back group operation step for system, different instant message applications.By taking wechat as an example, table 1 show wechat Symbian, Group operation step is moved back on Android, iphone and Windows phone.

1 wechat of table moves back group operation step on different operating system

OS Type	Move back group operation
		Symbian	Wechat-group-option-is deleted and is exited
Android	The more human head pictures in wechat-group-click upper right corner-upper right corner session operation-are deleted and are exited
		iphone	Wechat-group-more human head pictures-in the click upper right corner are deleted and are exited
Windows phone	The action bar-of the more human head pictures in wechat-group-click upper right corner-click bottom is deleted and is exited

Still by taking iphone mobile phones version " wechat " shown in Fig. 1 (a) as an example, if user want to exit out " party discuss " this Group, user needs to first click on " party discusses ", into this group, as shown in Fig. 1 (b), clicks the more human head pictures in the upper right corner, The chat message interface as shown in Fig. 1 (c) is popped up, " delete and exit " button is then clicked, then can be exited " party discusses " This group.

From the foregoing, it will be observed that existing group operation of moving back is required for manual operation, and need to execute the click behaviour of an at least step or more Make, if there are many interim group, cumbersome moves back group operation, and prodigious influence is caused to the experience of user.On the other hand, due to User actively moves back group, it will usually which the prompt message that " so-and-so has dropped out group " is shown on group chat interface can inevitably cause user to move back The embarrassment of group.

Thus, inventor by research, it is proposed that it is a kind of can automatic identification user sentence in the behavioural characteristic of some group Whether disconnected user needs to exit the group, and then in the case where judging that user needs to exit the group, prompts user to execute and really moves back Group recognizes operation, to realize that a key successfully moves back group, simplifies the troublesome operation for moving back group in traditional immediate communication tool by hand, improves User experience effect.

In a kind of optional application environment, Fig. 2 shows a kind of optional instant messaging network architecture schematic diagrams, such as scheme Shown in 2, server can be the application server for providing any one instant messaging service (for example, wechat, QQ etc.), visitor Family end (Fig. 2 shows N number of) can be to be equipped with instant messaging tools, or can be based on Web page and access Instant Messenger conviction The terminal device (for example, mobile phone, computer, laptop, tablet computer etc.) of business.

In a kind of scene, an instant messenger can be installed in each client shown in Fig. 2, this is immediately At least one session group (for example, wechat group) can be provided on chat tool (for example, wechat), belong to the multiple of same session group Client may be constructed an instant chat room, server can obtain in real time in the instant chat room each client this i.e. When chatroom in group session feature.After user logs in instant messenger (for example, wechat) in any one client, Client can start the group's state feature for detecting each session group in current instant messenger (for example, wechat) interface, and root According to the behavioural characteristic of user's each session group in current instant messenger interface, determine that user exits the general of each session group Rate, by taking wechat group list shown in Fig. 1 (a) as an example, client is in real time or timing detects behavior of the user in each session group Feature finds that this session group does not all have user within a very long time " party discusses " through detection after a period of time Any session content was carried out, and often there are other users to exit this session group, in this case, it can be determined that user It is possible that " party discusses " this group is wanted to exit out, at this point, client can prompt the user whether to exit " party discusses " group Group exits if user confirms, group operation is moved back in execution, moves back group operation conversely, not executing then, and record the operation of user.It can Selection of land can calculate each user according to user behavior characteristics of each user of collection in session group and exit the general of session group Rate, and whether the determine the probability for exiting according to each user session group prompts the user with and moves back group operation.A kind of optional scheme is such as Under：

For example, client 1 can initiate all users in the local at least one session group for working as front opening of traversal in Fig. 2 User characteristics behavior, and the user behavior characteristics for each user for including in each session group got are sent to service Device, similarly, server can also be received to be fed back from other clients (for example, client 2, client 3 ... client N) Each session group in all users user behavior characteristics, and in each session group from multiple client feedback user Behavioural characteristic carries out duplicate removal processing, and updates the spy of group's state feature of each session group in the model file stored on server Value indicative pre-saves the user behavior characteristics for the session group fed back from multiple client, updated mould in the model file Type file is also provided to each client shown in Fig. 2, determines that user exits each session group for each client Probability.

Herein it should be noted that the software platform of above-mentioned each client can be, but not limited to Symbian, Android, The operating system platforms such as iphone and Windows phone；Its hardware platform can be, but not limited to computer, laptop, The equipment such as mobile phone, tablet computer.

Under above application scene, Fig. 3 is shown optionally automatically exits from Instant Messenger according to one kind of the embodiment of the present application The system schematic for believing session group, as shown in figure 3, the system includes：Server 101 and client 103.

Wherein, server 101, for providing model file, wherein model file has pre-saved at least one client The user behavior characteristics of the session group of feedback, the user behavior characteristics of session group include the user of at least one user in session group Behavioural characteristic；

Client 103 is communicated with server 101, and for traversing at least one session group, include in acquisition session group is every The user behavior characteristics of a user, and according to the user behavior characteristics for each user for including in session group, obtain each user The probability for exiting session group, according to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.

It should be noted that above-mentioned client 103 can be in N number of client in the above-mentioned network architecture shown in Fig. 2 Any one client, the instant communication software installed in the client can include but is not limited to QQ, wechat, Fetion, MSN Deng；Above-mentioned server can be the application server for providing instant messaging service, for example, QQ servers, wechat server, Fetion Server, MSN servers etc..Client 103 can be with server 101 by internet communication, and obtains and deposited on server 101 The model file of storage, due at least one of the session group that contains at least one client feedback in model file user User behavior characteristics, client can by traverse installation instant communication software on the client or based on Web i.e. When communication server interface in each session group, obtain the user behavior characteristics for each user for including in session group, and root According to the user behavior characteristics for each user for including in session group, the probability that each user exits session group, last basis are obtained Each user exits the probability of session group, it is determined whether the prompt message of session group is exited to user output.

From the foregoing, it will be observed that in scheme disclosed in the above embodiments of the present application 1, client 103 is communicated with server 101, is passed through The model file that server 101 provides, the one or more in traverse instant communication software of client 103 and are got session groups In session group after the user behavior characteristics of each user, according to the user behavior characteristics for each user for including in session group, from The probability that each user in the session group exits the session group is obtained on server 101, and the meeting is then exited according to each user Talk about the probability of group, it is determined whether the prompt message of the session group is exited to user's output.

It is easily noted that, since client 103 can obtain the user behavior characteristics of each user in session group automatically, In the case that the session group of user includes multiple, executed by hand without user it is cumbersome move back group operation to exit some group, visitor Family end 103 receives user behavior characteristics of the server according to user in session group, and user is calculated and exits each session group Probability the prompt message of session group is exited to user's output and in the case where probability meets preset condition, with realize from It is dynamic to move back group.Therefore, the scheme provided by the embodiment of the present application, the user behavior reached according to user in session group are special It levies to predict that user exits the probability of the session group, to determine whether the purpose for prompting user's execution to move back group operation, to realize Simplified user executes and moves back group operation, improves the technique effect of user experience.

As a result, the scheme of above-described embodiment 1 provided by the present application solve existing instant messaging tools to move back group operation numerous The technical issues of trivial influence user experience effect.

In order to enable server 101 gets the user for each user for including in each session group in instant messenger Behavioural characteristic, in a kind of optional embodiment, what above-mentioned client 103 can be shown in real time or in timing traversal client end interface At least one session group, and the user that each session group is included is obtained, and detect and each used included in each session group At least one group's state feature at family obtains session if detecting that at least one group's state feature of user changes The user behavior characteristics of at least one of the group changed user of group's state feature, and by each session group in client end interface User behavior characteristics be sent on server 101.

Based on above-described embodiment, at least one group's shape of each user included in session group is detected in client 103 Before state feature, client 103 can open at least one session group of application software in client end interface, and collect each meeting Group's state feature of group is talked about, if detecting that the event that user exits occurs for any one session group, preservation is exited Group's state feature of the user of event obtains group's state feature in the user for not exiting session group；Wherein, session group people It is average that number, session group history chat messages number, session group participate in message in chat number, the establishment duration of session group, session group By at most chat messages number, session group in average everyone chat messages number, session group in reading number, session group by personnel in group Participation number and session group chat messages in the number of top set, the nearest chatting time of nearest chatting time, session group of session group Number.

Further, after the user behavior characteristics for each user for including in getting session group, client 103 can be with According to the model file stored on server 101, the probability that each user in session group exits session group is calculated, as one Kind optional embodiment, client 103 is from 101 reading model file of server, and according to the user recorded in model file Behavioural characteristic carries out probability calculation to the user behavior characteristics for each user for including in session group, obtains each user's meeting of exiting Talk about the probability of group, wherein model file has pre-saved the user behavior characteristics of the session group of at least one client feedback, meeting The user behavior characteristics of words group include the user behavior characteristics of at least one user in session group.

As a kind of optional embodiment, the session group pair can be obtained according to the variation of group's state feature of session group The user behavior characteristics of the user answered, specifically, client 103 obtain at least one group of states changed in session group After feature, in the group's state feature being locally stored before update session group variation, and the session that group's state feature is changed For at least one group of state Feature Mappings of group to server 101, server is receiving the session group that is changed at least After one group's state feature, group's state feature before session group changes in server 101 is changed, and according to updated group State feature updates the user behavior characteristics of at least one of each session group pre-saved in model file user.

In the above-described embodiments, the event of exiting has occurred in preservation in embodiment as one preferred, client 103 Group's state feature of user after obtaining group's state feature of each session group in the open state, can also will be in and beat Group's state feature of each session group of open state is transmitted to server after carrying out compression processing, and specifically, client 103 can be with Classify to multiple session groups according to the event of exiting whether is generated in session group in the open state, obtains first kind sample With the second class sample, and calculate the Euclidean distance in first kind sample between any two group state feature, and select it is European away from The first compression processing is carried out from two group's state features less than the first predetermined threshold, obtains compressed third class sample, together When, the Euclidean distance between any two group's state feature in the second class sample is calculated, and it is pre- to select Euclidean distance to be less than second Two group's state features for determining threshold value carry out the second compression processing, obtain compressed 4th class sample；It finally will be by third class Sample and the 4th class sample delivery are to server；Wherein, first kind sample, which contains in the open state and user and sends out, exits Group's state feature of session group, the second class sample are group's state feature in the open state and not exiting session group.

Embodiment as one preferred, multiple client 103 by third class sample and the 4th class sample delivery extremely After server 101, server 101 receives third class sample and the 4th class sample from multiple client, obtains multiple Third class sample and multiple 4th class samples, and calculate in multiple third class samples between any two characteristic it is European away from From select Euclidean distance to be less than in two characteristics of third predetermined threshold one is retained, and calculates the 4th class sample Euclidean distance in this between any two characteristic selects Euclidean distance to be less than two characteristics of the 4th predetermined threshold In one retained；Statistics is retained in the characteristic value of sample data in multiple third class samples, and count be retained in it is more The characteristic value of sample data in a 4th class sample generates model file.

It should be noted that characteristic value is used to characterize going out for group's state feature of each session group of multiple client return Existing frequency, occurrence number.

Optionally, server 101 is retained in the characteristic value of the sample data in multiple third class samples in statistics, and counts It is retained in the characteristic value of the sample data in multiple 4th class samples, before generating model file, can also will be retained in multiple Sample data in third class sample is respectively mapped to scheduled data with the sample data being retained in multiple 4th class samples Range, and the label within the scope of flag data.

In a kind of optional embodiment, each user can be calculated using bayesian algorithm and exits the general of session group Rate, specifically, server 101 can be obtained according to the characteristic value of the user behavior characteristics of the session group recorded in model file The characteristic value of the user behavior characteristics of each user shown in session window, and using bayesian algorithm to being shown in session window The characteristic value of the user behavior characteristics of each user shown carries out calculation processing, obtains each user and exits the general of each session group Rate.

Specifically, server 101 using bayesian algorithm to client end interface in group's shape of each session group for showing The characteristic value of state feature carries out calculation processing, and during the probability for obtaining opening each session group, server 101 can be distinguished The characteristic value of group's state feature of each session group shown in client end interface obtains the spy of multiple first monoid state features The characteristic value of value indicative and multiple second monoid state features, using bayesian algorithm respectively to multiple first monoid state features The characteristic value of characteristic value and multiple second monoid state features carry out calculation processing, obtain each session group it is in the open state and The first probability of exit message is generated, and each session group is in the open state and does not generate the second probability of exit message； Wherein, session group is in the open state and generation exit message, the second monoid state are special for characterizing for the first monoid state feature It takes over for use in the open state in characterization session group and does not generate exit message.

In a kind of optional embodiment, above-mentioned server 101 can be also used for according to each session group corresponding first Probability and the second probability, there are the priority that exits in each session group of same subscriber for calculating, and foundation exits priority pair Each session group is ranked up, and obtains exiting exiting sequentially for each session group.

Embodiment 2

According to the embodiment of the present application, a kind of embodiment of the method automatically exiting from instant communication session group, this reality are additionally provided The method for automatically exiting from instant communication session group for applying example offer can be applied to automatically exiting from i.e. for the offer of the embodiment of the present application 1 When communication session group system in, including but not limited to the embodiment of the present application 1 describe application scenarios.It should be noted that The step of flow of attached drawing illustrates can execute in the computer system of such as a group of computer-executable instructions, also, It, in some cases, can be with different from shown in sequence execution herein although logical order is shown in flow charts The step of going out or describing.

Since existing group operation of moving back is required for manual operation, and need to execute the clicking operation of an at least step or more, such as If there are many interim group of fruit, cumbersome moves back group operation, and prodigious influence is caused to the experience of user.It is provided in this embodiment automatic The use of each session group in automatic identification user user's instant messaging tools may be implemented in the method for exiting instant communication session group Family behavioural characteristic judges that user exits the probability of each session group, and then exits the determine the probability of each session group according to user Whether prompt user's execution really to move back group and recognize operation, user only needs execution one click confirmation operation that can realize that a key successfully moves back Group, simplifies the troublesome operation for moving back group in traditional immediate communication tool by hand, improves user experience effect.

Fig. 4 is according to a kind of flow chart of method automatically exiting from instant communication session group of the embodiment of the present application, such as Fig. 4 It is shown, include the following steps：

Step S402 traverses at least one session group, obtains the user behavior characteristics for each user for including in session group.

Specifically, in above-mentioned steps, above-mentioned session group can be created in instant messaging tools comprising multiple users Group, can support more human hair sending voices, video, picture or word conversate chat, for example, QQ groups, QQ discussion group, Wechat group etc.；Above-mentioned user behavior characteristics can be used for characterizing each behavioural characteristic of the user in the session group in session group, For example, actively initiating the number of session, the accumulative time to conversate, the number etc. by other [email protected] instant message applications In in real time or timing traverses one or more session groups, the user behavior for obtaining in each session group each user for including is special Sign.

Optionally, above-mentioned instant messaging tools can be but not limited to the immediate communication tools such as QQ, wechat, Fetion, MSN.

Step S404 obtains each user's meeting of exiting according to the user behavior characteristics for each user for including in session group Talk about the probability of group.

Specifically, in above-mentioned steps, the above-mentioned probability for exiting session group can be system according to the user of collection in meeting It talks about in group after a large amount of behavioural characteristic, the probability of the session group may be exited by obtaining user by certain mathematical model；? After the user behavior characteristics for getting each user for including in session group, as a kind of optional embodiment, it can transfer Preset model, the probability of the session group may be exited by obtaining each user.

Step S406, according to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.

Specifically, in above-mentioned steps, above-mentioned prompt message can be to move back group confirmation operation for prompting user to execute Information obtains after it exits the probability of the session group, it can be determined that should when the user behavior characteristics according to user in session group Whether the probability that user exits the session group meets preset condition, if meeting preset condition, exports prompt user's meeting of exiting Talk about the prompt message of group；Conversely, not making any operation then.

Herein it should be noted that being moved back at least one of session group user behavior characteristics to obtain user according to user Go out in the probabilistic process of the session group, the user behavior characteristics of consideration are more, get each user in session group and exit the meeting The probability for talking about group is more accurate.

In a kind of optional embodiment, by taking " wechat " of mobile phone version as an example, Fig. 1 (a) is shown according to the embodiment of the present application " wechat " group chat list schematic diagram of mobile phone version a kind of 6 session groups are contained in the group chat interface as shown in Fig. 1 (a), Respectively " product department ", " party discusses ", " corporate HQ ", " senior middle school classmate ", " red packet group " and " mobile phone purchase exchanges ".Assuming that " mobile phone purchase exchange " group is that current (on March 20th, 2017) user (on October 1st, 2016) when buying mobile phone is added One group, and some gains in depth of comprehension of purchase mobile phone are linked up and shared with good friend, later never again in the session group Session was carried out, based on scheme disclosed in above-mentioned steps S202 to S206, active user can be predicted and exit the session group Probability is larger, thus can export the prompt message for prompting the user whether to exit the session group.And for " senior middle school classmate " this Session group, it is assumed that the time that user creates this group is (on September 1st, 2000), and nearly year, user is not at this Session group carried out session, but detected that the group members of more than half are all friend relations in user and the session group, or warp Often@can be carried out by the group members of other in session group, although the then all no meeting of the user nearly year in the session group Words behavior is predicted active user and is exited the probability of the session group also not based on scheme disclosed in above-mentioned steps S202 to S206 Can be too big, because without prompting user to exit the prompt message of the session group.

As a kind of optional embodiment, when each user exits the probability satisfaction of session group and exits session in session group After the probability of group, the prompt message that session group is exited to user's output can be the information for allowing user to execute a key confirmation operation, If the user desired that exit the session group, then only need to click " confirmation " button shown on interface, and may be implemented a key at Work(moves back the purpose of group；If user is not intended to exit the session group, the " Cancel " button shown on interface can be clicked, to Group operation is moved back in cancellation.

From the foregoing, it will be observed that in scheme disclosed in the above embodiments of the present application 2, by traversing in instant communication software one or more A session group, the user behavior characteristics of each user in real time or in timing acquisition session group, and it is every according to include in session group The user behavior characteristics of a user obtain the probability that each user in the session group exits the session group, are then used according to each The probability of the session group is exited at family, it is determined whether the prompt message of the session group is exited to user's output.

It is easily noted that, since the user behavior characteristics of each user in session group obtain automatically, in the meeting of user Talk about in the case that group includes multiple, executed by hand without user it is cumbersome move back group operation to exit some group, system can root According to user behavior characteristics of the user obtained automatically in session group, goes out user from multiple session group automatic Predictions and may wish to move back Group moves back automatically to realize in the group gone out.Therefore, the scheme provided by the embodiment of the present application, has reached according to user in meeting The user behavior characteristics in group are talked about to predict that user exits the probability of the session group, group behaviour is moved back to determine whether to prompt user to execute The purpose of work moves back group operation to realize simplified user's execution, improves the technique effect of user experience.

As a result, the scheme of above-described embodiment 2 provided by the present application solve existing instant messaging tools to move back group operation numerous The technical issues of trivial influence user experience effect.

In a kind of optional embodiment, as shown in figure 5, traversing at least one session group, include in acquisition session group The user behavior characteristics of each user, may include steps of：

Step S502, at least one session group shown in real time or in timing traversal client end interface, obtains any one The user that session group is included, wherein session group corresponds to a session window；

Step S504 detects at least one group's state feature of each user included in session group；

Step S506 is obtained in session group extremely if detecting that at least one group's state feature of user changes The user behavior characteristics of a few changed user of Population status feature.

Specifically, in above-mentioned steps, client can be equipped with for instant communication software (for example, QQ, wechat, Fetion, MSN etc.) the smart machines such as computer, laptop, tablet computer, mobile phone；Above-mentioned interface can be in client The chat interface of the instant communication software of installation, or Web editions instant messaging chat interfaces are based on, it is micro- by taking wechat as an example Letter is used as a instant communication software, there is three kinds of mobile phone version, computerized version and Web versions forms.Above-mentioned group's state feature can be i.e. When communication software in for characterizing the feature of each session group state, in a kind of optional embodiment, by taking " wechat " as an example, visitor Group's state feature of each session group can include but is not limited in the interface of family end：Session group number, the chat of session group history disappear It is average by reading number, session group to cease message in number, session group participation chat number, the establishment duration of session group, session group At most chat messages numbers, session group is by the number of personnel's top set in group, session group in average everyone chat messages number, session group Participation number and session group chat messages number in nearest chatting time, the nearest chatting time of session group.It should be noted that each Session group can correspond to a session window, that is, carry out the session window of group chat.

In a kind of optional embodiment, above-mentioned group's state feature can include but is not limited to feature shown in table 2.

2 groups of state features of table

By scheme disclosed in above-described embodiment, realizes and obtained in session group often according to session group state characteristic The purpose of the user behavior characteristics of a user.

Based on above-described embodiment, in the user for getting the changed user of at least one of session group group's state feature After behavioural characteristic, in a kind of optional embodiment, as shown in fig. 6, according to the user behavior for each user for including in session group Feature obtains the probability that each user exits session group, may include steps of：

Step S602, reading model file, wherein model file has pre-saved the session of at least one client feedback The user behavior characteristics of group, the user behavior characteristics of session group include the user behavior characteristics of at least one user in session group；

Step S604, according to the user behavior characteristics recorded in model file to the use for each user for including in session group Family behavioural characteristic carries out probability calculation, obtains the probability that each user exits session group.

Specifically, in above-mentioned steps, above-mentioned model file can be that each client that contains that server provides is led to Cross instant communication software carry out group session each session group group's state feature and these group of state feature characteristic value File；Characteristic value can be the frequency that group's state feature of each session group of at least one user end to server feedback occurs Deng.One or more group's states of each session group shown in real time or in timing detection client instant communication software interface are special During sign, if detecting that at least one group of state features of any one session group in current interface change, The reading model file from server, and it is special to the user behavior for each user for including in each session group according to model file Sign carries out probability calculation, and the probability of the session group may be exited by obtaining each user.

As an alternative embodiment, the probability for exiting session group may include two kinds, one is prompt users to move back User confirms the probability for exiting session group after going out session group, i.e., is just exiting the probability of session group；One is prompt users to exit meeting User does not exit the probability of session group after words group, i.e., the negative probability for exiting session group.

In a kind of optional embodiment, as shown in fig. 7, if detecting that at least one group's state feature of user occurs Variation, then obtain the user behavior characteristics of the changed user of at least one of session group group's state feature, may include as Lower step：

Step S702 obtains at least one group of state features changed in session group；

Step S704, in the group's state feature being locally stored before update session group variation；

At least one group of state Feature Mappings of step S706, the session group that group's state feature is changed extremely service Device changes group's state feature before session group changes in server；

Step S708 is updated according to updated group's state feature in each session group pre-saved in model file At least one user user behavior characteristics.

Specifically, in above-mentioned steps, above-mentioned server can be used for storing chat circle for containing instant communication software Group's state feature of each session group in face and the model file of its characteristic value；It is shown in detecting client end interface In the case that variation has occurred at least one group's state feature of each session group, client can be become according in session group At least one group of state features changed, the group's state feature for the session group being locally stored before update session group variation, in client When timing or client closing are exited, at least one group of state features of the session group that group's state feature is changed are reflected Be incident upon server, server after receiving at least one group of state features of the session group that group state feature is changed, Group's state feature before session group changes in modification server, and according to updated group's state feature, more new model text The characteristic value of each user behavior characteristics in the session group pre-saved in part, wherein characteristic value includes one of following：Group's state The frequency of occurrences, the occurrence number of feature.

By above-described embodiment, realize that server end is real-time or the group of each session group of each client of timing acquisition State feature, and update the purpose of model file.

It updates and is protected in advance in model file on server according to updated group's state feature based on above-described embodiment After the user behavior characteristics of at least one of session group deposited user, in a kind of optional embodiment, as shown in figure 8, according to The user behavior characteristics value recorded in model file carries out probability to the user behavior characteristics for each user for including in session group It calculates, obtains the probability that each user exits session group, may include steps of：

Step S802 obtains conversation window according to the characteristic value of the user behavior characteristics of the session group recorded in model file The characteristic value of the user behavior characteristics of each user shown in mouthful；

Step S804, using bayesian algorithm to the feature of the user behavior characteristics of each user shown in session window Value carries out calculation processing, obtains the probability that each user exits each session group.

Specifically, in above-mentioned steps, due to the model file stored on server have recorded it is each in each client The characteristic value of the user behavior characteristics and each user behavior characteristics of each user in session group, thus, utilizing model text The user behavior characteristics recorded in part carry out the user behavior characteristics for each user for including in session group the mistake of probability calculation Cheng Zhong can obtain group session according to the characteristic value of the user behavior characteristics recorded from the model file read on server The characteristic value of the user behavior characteristics of each user shown in window, and using bayesian algorithm to showing in session window The characteristic value of the user behavior characteristics of each user carries out calculation processing, obtains the probability that each user exits each session group.

In a kind of optional embodiment, as shown in figure 9, at least one of each user included in detection session group Before Population status feature, the above method can also include the following steps：

Step S902, opens at least one session group of application software in client end interface, and collects each session group Group's state feature；

Step S904, if detecting that the event that user exits occurs for any one session group, preservation is exited Group's state feature of the user of event obtains group's state feature in the user for not exiting session group；

Wherein, group's state feature includes at least one of following：Session group number, session group history chat messages number, session It is average by averagely everyone chats in reading number, session group that group participates in message in chat number, the establishment duration of session group, session group Most chat messages numbers in its message count, session group, session group by the number of personnel's top set in group, session group nearest chat when Between, participation number and session group chat messages number in the nearest chatting time of session group.

Specifically, in above-mentioned steps, in the local of each client, there are one characteristic storage modules, for storing visitor Group's state feature of each session group in the chat interface of family end, when user opened in client end interface one of application software or After multiple session groups, group's state feature of the client meeting each session group of real-time collecting, and appoint in detecting client end interface In the case that dialog events occur for one session group of meaning, the session that the event of exiting has occurred is preserved in local characteristic storage module Group's state feature of group, obtains group's state feature of each session group in the open state.

By above-described embodiment, the purpose for the group's state feature for obtaining each session group of client in real time is realized.

It is as shown in Figure 10, special in the group's state for preserving the user that the event of exiting has occurred in a kind of optional embodiment Sign, after obtaining group's state feature of each session group in the open state, the above method can also include the following steps：

Step S102 is transmitted to clothes after group's state feature of each session group in the open state is carried out compression processing Business device, the step include：

Step S102a carries out multiple session groups according to the event of exiting whether is generated in session group in the open state Classification, obtains first kind sample and the second class sample, wherein first kind sample, which contains in the open state and user and sends out, to move back Go out group's state feature of session group, the second class sample is group's state feature in the open state and not exiting session group；

Step S102b calculates the Euclidean distance between any two group's state feature in first kind sample, and selects European Two group's state features that distance is less than the first predetermined threshold carry out the first compression processing, obtain compressed third class sample；

Step S102c calculates the Euclidean distance between any two group's state feature in the second class sample, and selects European Two group's state features that distance is less than the second predetermined threshold carry out the second compression processing, obtain compressed 4th class sample；

Step S102d, by third class sample and the 4th class sample delivery to server.

Specifically, in above-mentioned steps, after group's state feature that client preserves the session group that dialog events have occurred, Client is transmitted to server after group's state feature of each session group in the open state being carried out compression processing.It is a kind of In optional embodiment, whether can be executed according to user's opening session group and exit session group operation come to each session group Group's state feature obtains that session group is in the open state and user executes the first kind sample for exiting session group operation according to classifying Originally and session group is in the open state and user is not carried out the second class sample for exiting session group operation.For first kind sample Originally with the second class sample, the Euclidean distance between arbitrary two group characters is calculated separately, for arbitrary two group characters apart from small In predetermined threshold, then a group character is randomly choosed, to realize group compression of state characteristic, by first kind sample After the second class sample compression, third class sample and the 4th class sample are respectively obtained, third class sample and the 4th class sample are passed Transport to server.

It should be noted that by taking first kind sample as an example, it is assumed that include n session group, then each group in first kind sample State feature all corresponds to n session group, and thus, each group's state feature has a n characteristic value, two group's state features (A and B) it Between the formula that can be expressed as of distance：

Wherein, x_AkIndicate the characteristic value of group's state feature A of k-th of session group, x_BkIndicate group's shape of k-th of session group The characteristic value of state feature B；Group state feature A and group's state feature B is two n-dimensional vector a (x_A1, x_A2... x_An) and b (x_B1, x_B2... x_Bn)。

By above-described embodiment, group's state characteristic of client is subjected to compression and is retransmited to server, is improved Transmission rate reduces network flow cost.

In a kind of optional embodiment, as shown in figure 11, extremely serviced by third class sample and the 4th class sample delivery After device, the above method can also include the following steps：

Step S112, server receive third class sample and the 4th class sample from multiple client, obtain multiple Third class sample and multiple 4th class samples；

Step S114 calculates the Euclidean distance between any two characteristic in multiple third class samples, selects European Distance is retained less than one in two characteristics of third predetermined threshold, and calculates any two in the 4th class sample Euclidean distance between characteristic, a progress for selecting Euclidean distance to be less than in two characteristics of the 4th predetermined threshold Retain；

Step S116, statistics are retained in the characteristic value of sample data in multiple third class samples, and count be retained in it is more The characteristic value of sample data in a 4th class sample generates model file.

Specifically, in above-mentioned steps, server meeting timing acquisition third class sample and the 4th class sample are a kind of optional In embodiment, timing can be one day, and server receives multiple third class sample datas from multiple client and more A 4th class sample data is calculated separately for the data in the data and multiple 4th class samples in multiple third class samples Euclidean distance between arbitrary two groups state characteristic is less than the distance of arbitrary two groups state characteristic predetermined Threshold value then randomly chooses a certain item group state feature and retains, and counts third class sample and the 4th class sample upon compression The characteristic values such as the frequency of occurrences of group's state feature or occurrence number in this, it is raw according to its corresponding characteristic value of group's state feature At model file, storage in the server, is updated for client timing.

By above-described embodiment, group state characteristic of the server end from each client is subjected to compression processing, Realize the purpose for saving memory space.

In a kind of optional embodiment, characteristic value is used to characterize group's state of each session group of multiple client return The frequency of occurrences, the occurrence number of feature, wherein it is retained in the characteristic value of the sample data in multiple third class samples in statistics, And the characteristic value for being retained in the sample data in multiple 4th class samples is counted, before generating model file, as shown in figure 12, on The method of stating can also include the following steps：

Step S122 by the sample data being retained in multiple third class samples and is retained in multiple 4th class samples Sample data is respectively mapped to scheduled data area, and the label within the scope of flag data.

Specifically, in above-mentioned steps, due to being all data statistics value, the various increasing of number of statistical value in group's state feature Add the complexity of off-line training, thus, in a kind of optional embodiment, data can be mapped, define data area It is 0 to MAX, is divided into 10, that is, has MAX/10 interval number, interval mark, will be in group's state feature from 0 to (MAX/10-1) Characteristic value is replaced by interval mark.

In a kind of optional embodiment, Figure 13 is to map schematic diagram according to a kind of optional data of the embodiment of the present application； As shown in figure 13, if MAX is 20, in the case of being divided into 10, interval number 2, interval mark is respectively 0,1, then will be located at The value of 0-9 is replaced with 0, will be replaced with 1 positioned at the value of 10-19.

By above-described embodiment, the complexity for reducing off-line training is realized.

It is as shown in figure 14, every to what is shown in client end interface using bayesian algorithm in a kind of optional embodiment The characteristic value of group's state feature of a session group carries out calculation processing, obtains the probability for opening each session group, may include as Lower step：

Step S142 distinguishes the characteristic value of group's state feature of each session group shown in client end interface, obtains more The characteristic value of the characteristic value of a first monoid state feature and multiple second monoid state features, wherein the first monoid state is special It takes over for use in the open state in characterization session group and generates exit message, the second monoid state feature is in for characterizing session group Opening state and exit message is not generated；

Step S144 uses the bayesian algorithm characteristic value to multiple first monoid state features and multiple second classes respectively Group state feature characteristic value carry out calculation processing, obtain each session group it is in the open state and generate exit message first Probability, and each session group is in the open state and does not generate the second probability of exit message.

Specifically, in above-mentioned steps, in client according to the group's state recorded in the model file returned in server The characteristic value of feature, after obtaining the characteristic value of group's state feature of each session group shown in client end interface, by client The feature value division of group's state feature of each session group shown in interface is that session group is in the open state and generation is exited The characteristic value and session group of first monoid state feature of message are in the open state and do not generate the second monoid of exit message The characteristic value of state feature utilizes the bayesian algorithm characteristic value to multiple first monoid state features and multiple second classes respectively Group state feature characteristic value carry out calculation processing, obtain each session group it is in the open state and generate exit message first Probability, and each session group is in the open state and does not generate the second probability of exit message.

Client chat is obtained by calculation in the model file returned according to server by above-described embodiment, client The probability that each session group is opened by user in interface.

In a kind of optional embodiment, as shown in figure 15, the probability of each session group is exited to institute according to each user Some session groups are ranked up, and the sequence that exits for obtaining exiting each session group may include steps of：

Step S152, according to corresponding first probability of each session group and the second probability, there are the every of same subscriber for calculating Priority is exited in a session group；

Step S154 is ranked up each session group according to priority is exited, obtains exiting exiting for each session group Sequentially.

Specifically, in above-mentioned steps, each session is calculated according to model file, using bayesian algorithm in client After the first probability and the second probability of group, it is suitable can to obtain the processed priority of each session group by certain algorithm Sequence, and be ranked up session group in client chat interface according to the priority of each session group, it obtains exiting each session Group's exits sequence.

By above-described embodiment, the chat behavior using a large number of users and big data advantage are realized, it is soft to instant chat Session priority in part is ranked up, and improves the intelligent of chat software.

Embodiment as one preferred, Figure 16 is shown optionally to be automatically exited from i.e. according to one kind of the embodiment of the present application When communication session group overall procedure schematic diagram, as shown in figure 16, which mainly wraps Include two parts, offline feature data collection training module and on-line prediction module.Wherein, offline feature data collection trains mould Block is mainly responsible for from client and collects user behavior characteristics, carries out offline feature storage and training, on-line prediction module are mainly born Duty periodically traverses the group of all users, and is predicted according to the UserFeature of its group, when user needs to exit this group, User is then notified to carry out moving back group operation confirmation.

It can in the case that collection is characterized as user behavior characteristics shown in table 1 as a kind of optional embodiment It is as follows to define user behavior characteristics:

We define Datai={ Flagi, UserFeaturei }, and Datai indicates i-th record, the wherein value of Flag It is that 0 or 1,1 expression user moves back group by hand, 0 expression user indicates the user in this group without group operation, UserFeature is moved back Feature.When user exits by hand every time, server will record this behavior, data Data be stored, and user is moved back group's Data.Flag is set to 1.Since training pattern needs positive negative sample, we randomly choose the active group of user, count it UserFeature, and Flag is set to 0.

Training stage, we carry out data training using SVM, and the model file that training generates is synchronized to real-time prediction Module.

As a kind of optional embodiment, real-time online forecast function can be realized by following code：

By scheme disclosed in the above embodiments of the present application, group operation progress is moved back in advance to user using group behavior of moving back of user It surveys, carries out moving back group operation for user automatically.It avoids user's manual operations from moving back group, keeps simple interface.For the every of single user A group of behaviors carry out group based on algorithm and act on prediction, and judge whether user needs to exit the group.The advantages of the application is：1, It does not need user and moves back group by hand, there is great convenience；2, it carries out moving back group's judgement for single user, does not influence other use in group Family.

Optionally, as a kind of embodiment of expansion, other graders can be used to substitute SVM.

It should be noted that for each method embodiment above-mentioned, for simple description, therefore it is all expressed as a series of Combination of actions, but those skilled in the art should understand that, the application is not limited by the described action sequence because According to the application, certain steps can be performed in other orders or simultaneously.Secondly, those skilled in the art should also know It knows, embodiment described in this description belongs to preferred embodiment, involved action and module not necessarily the application It is necessary.

Through the above description of the embodiments, those skilled in the art can be understood that according to above-mentioned implementation The method for automatically exiting from instant communication session group of example can add the mode of required general hardware platform to realize by software, when So can also be by hardware, but the former is more preferably embodiment in many cases.Based on this understanding, the technology of the application Substantially the part that contributes to existing technology can be expressed in the form of software products scheme in other words, the computer Software product is stored in a storage medium (such as ROM/RAM, magnetic disc, CD), including some instructions were used so that an end The method that end equipment (can be mobile phone, computer, server or the network equipment etc.) executes each embodiment of the application.

Embodiment 3

According to the embodiment of the present application, additionally provide a kind of for realizing the above-mentioned method for automatically exiting from instant communication session group Device embodiment, Figure 17 is a kind of schematic device automatically exiting from instant communication session group according to the embodiment of the present application, As shown in figure 17, which includes：First acquisition module 171, the second acquisition module 173 and determining module 175.

Wherein, the first acquisition module 171 obtains each use for including in session group for traversing at least one session group The user behavior characteristics at family；

Second acquisition module 173 obtains each for the user behavior characteristics according to each user for including in session group User exits the probability of session group；

Determining module 175, for according to the probability for exiting session group, it is determined whether the prompt letter of session group is exited in output Breath.

Herein it should be noted that above-mentioned first acquisition module 171, the second acquisition module 173 and determining module 175 correspond to Step S402 to S406 in embodiment 2, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but It is not limited to the above embodiments 2 disclosure of that.It should be noted that above-mentioned module can be such as a part of of device It is executed in the computer system of a group of computer-executable instructions.

From the foregoing, it will be observed that in scheme disclosed in the above embodiments of the present application 3, Instant Messenger is traversed by the first acquisition module 171 Believe one or more session groups in software, the user behavior characteristics of each user in real time or in timing acquisition session group, and pass through Second acquisition module 173 is obtained and is each used in the session group according to the user behavior characteristics for each user for including in session group The probability of the session group is exited at family, and last determining module 175 exits the probability of the session group according to each user, it is determined whether to User exports the prompt message for exiting the session group.

It is easily noted that, since the user behavior characteristics of each user in session group obtain automatically, in the meeting of user Talk about in the case that group includes multiple, executed by hand without user it is cumbersome move back group operation to exit some group, first obtains mould User behavior characteristics of the user that block 171 can obtain automatically in session group, the second acquisition module 173 can be obtained according to first User behavior characteristics of the user that modulus block 171 is got in session group get the probability that user exits each session group, So that determining module 175 can exit the probability of each session group according to user, it is determined whether exit the session to user's output The prompt message of group.Therefore, the scheme provided by the embodiment of the present application has reached the user in session group according to user Behavioural characteristic predicts that user exits the probability of the session group, and the purpose of group operation is moved back to determine whether to prompt user to execute, from And realize simplified user's execution and move back group operation, improve the technique effect of user experience.

As a result, the scheme of above-described embodiment 3 provided by the present application solve existing instant messaging tools to move back group operation numerous The technical issues of trivial influence user experience effect.

In a kind of optional embodiment, as shown in figure 17, above-mentioned first acquisition module includes：Third acquisition module is used In at least one session group shown in real time or in timing traversal client end interface, the use that any one session group is included is obtained Family, wherein session group corresponds to a session window；Detection module, for detecting each user included in session group extremely A few Population status feature；4th acquisition module, if for detecting that at least one group's state feature of user changes, Then obtain the user behavior characteristics of the changed user of at least one of session group group's state feature.

Herein it should be noted that above-mentioned third acquisition module, detection module and the 4th acquisition module correspond to embodiment 2 In step S502 to S506, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but is not limited to State 2 disclosure of that of embodiment.It should be noted that above-mentioned module can be in such as one group of calculating as a part of of device It is executed in the computer system of machine executable instruction.

In a kind of optional embodiment, above-mentioned second acquisition module includes：Read module is used for reading model file, Wherein, model file has pre-saved the user behavior characteristics of the session group of at least one client feedback, the user of session group Behavioural characteristic includes the user behavior characteristics of at least one user in session group；First computing module, for according to model file The user behavior characteristics of middle record carry out probability calculation to the user behavior characteristics for each user for including in session group, obtain every A user exits the probability of session group.

Herein it should be noted that above-mentioned read module and the first computing module correspond to the step S602 in embodiment 2 To S604, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but it is public to be not limited to the above embodiments 2 institutes The content opened.It should be noted that above-mentioned module can be in such as a group of computer-executable instructions as a part of of device Computer system in execute.

In a kind of optional embodiment, above-mentioned 4th acquisition module includes：5th acquisition module, for obtaining session group In at least one group of state features being changed；First update module, for being locally stored before updating session group variation Group's state feature；Modified module, at least one group of state features of the session group for being changed group's state feature Server is mapped to, group's state feature before session group changes in server is changed；Second update module, for according to more Group's state feature after new updates the user behavior of at least one of each session group pre-saved in model file user Feature.

Herein it should be noted that above-mentioned 5th acquisition module, the first update module, modified module and the second update module Corresponding to the step S702 to S708 in embodiment 2, example and application scenarios phase that above-mentioned module and corresponding step are realized Together, but 2 disclosure of that are not limited to the above embodiments.It should be noted that above-mentioned module can be with as a part for device It is executed in the computer system of such as a group of computer-executable instructions.

In a kind of optional embodiment, above-mentioned first computing module includes：6th acquisition module, for according to model text The characteristic value of the user behavior characteristics of the session group recorded in part obtains the user behavior of each user shown in session window The characteristic value of feature；Second computing module, for the user using bayesian algorithm to each user shown in session window The characteristic value of behavioural characteristic carries out calculation processing, obtains the probability that each user exits each session group.

Herein it should be noted that above-mentioned 6th acquisition module and the second computing module correspond to the step in embodiment 2 S802 to S804, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but is not limited to the above embodiments 2 Disclosure of that.It should be noted that above-mentioned module can be executable in such as one group of computer as a part of of device It is executed in the computer system of instruction.

In a kind of optional embodiment, above-mentioned apparatus further includes：Collection module is answered for being opened in client end interface With at least one session group of software, and collect group's state feature of each session group；Memory module, if taken office for detecting The event that user exits occurs for one session group of meaning, then preserves the group's state feature for the user that the event of exiting has occurred, obtain everywhere In the group's state feature for the user for not exiting session group；Wherein, session group number, session group history chat messages number, session group It is average by everyone average chat in reading number, session group to participate in message in chat number, the establishment duration of session group, session group Most chat messages numbers in message count, session group, session group by the number of personnel's top set in group, session group nearest chat when Between, participation number and session group chat messages number in the nearest chatting time of session group.

Herein it should be noted that above-mentioned collection module and memory module correspond to the step S902 in embodiment 2 extremely S904, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but is not limited to the above embodiments disclosed in 2 Content.It should be noted that above-mentioned module can be in such as a group of computer-executable instructions as a part of of device It is executed in computer system.

In a kind of optional embodiment, above-mentioned apparatus further includes：Processing module, being used for will be each of in the open state Group's state feature of session group is transmitted to server after carrying out compression processing, which includes：Sort module is used for basis The event of exiting whether is generated in session group in the open state to classify to multiple session groups, obtains first kind sample and Two class samples, wherein first kind sample contains in the open state and user and sends out the group's state feature for exiting session group, the Two class samples are group's state feature in the open state and not exiting session group；Third computing module, for calculating first Euclidean distance in class sample between any two group state feature, and select Euclidean distance less than two of the first predetermined threshold Group's state feature carries out the first compression processing, obtains compressed third class sample；4th computing module, for calculating the second class Euclidean distance in sample between any two group state feature, and select two groups of the Euclidean distance less than the second predetermined threshold State feature carries out the second compression processing, obtains compressed 4th class sample；Transmission module, for by third class sample and the Four class sample deliveries are to server.

Herein it should be noted that above-mentioned processing module, sort module, third computing module, the 4th computing module and biography The example that step S102, S102a that defeated module corresponds in embodiment 2 is realized to S102d, above-mentioned module with corresponding step It is identical with application scenarios, but it is not limited to the above embodiments 2 disclosure of that.It should be noted that above-mentioned module is as device A part of can be executed in the computer system of such as a group of computer-executable instructions.

In a kind of optional embodiment, above-mentioned apparatus further includes：Receiving module is received for server from multiple The third class sample and the 4th class sample of client, obtain multiple third class samples and multiple 4th class samples；5th calculates mould Block selects Euclidean distance to be less than for calculating the Euclidean distance in multiple third class samples between any two characteristic One in two characteristics of three predetermined thresholds is retained, and calculate the 4th class sample in any two characteristic it Between Euclidean distance, select Euclidean distance be less than the 4th predetermined threshold two characteristics in one retained；Statistics Module, the characteristic value for counting the sample data being retained in multiple third class samples, and count and be retained in multiple 4th classes The characteristic value of sample data in sample generates model file.

Herein it should be noted that above-mentioned receiving module, the 5th computing module and statistical module correspond in embodiment 2 Step S112 to S116, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but is not limited to above-mentioned reality Apply 2 disclosure of that of example.It should be noted that above-mentioned module can be such as one group of computer can as a part of of device It is executed in the computer system executed instruction.

In a kind of optional embodiment, characteristic value is used to characterize group's state of each session group of multiple client return The frequency of occurrences, the occurrence number of feature, wherein above-mentioned apparatus method further includes：Mapping block, for multiple thirds will to be retained in Sample data in class sample is respectively mapped to scheduled data area with the sample data being retained in multiple 4th class samples, And the label within the scope of flag data.

Herein it should be noted that above-mentioned mapping block correspond to embodiment 2 in step S122, above-mentioned module with it is corresponding The step of the example realized it is identical with application scenarios, but be not limited to the above embodiments 2 disclosure of that.It needs to illustrate It is that above-mentioned module can be executed as a part of of device in the computer system of such as a group of computer-executable instructions.

In a kind of optional embodiment, the second computing module includes：Division module is shown for distinguishing in client end interface The characteristic value of the group's state feature for each session group shown obtains the characteristic value and multiple second of multiple first monoid state features The characteristic value of monoid state feature, wherein session group is in the open state and generation is moved back for characterizing for the first monoid state feature Outbound message, the second monoid state feature are in the open state and do not generate exit message for characterizing session group；6th calculates mould Block, for using the bayesian algorithm characteristic value to multiple first monoid state features and multiple second monoid state features respectively Characteristic value carry out calculation processing, obtain each session group it is in the open state and generate exit message the first probability, and Each session group is in the open state and does not generate the second probability of exit message.

Herein it should be noted that above-mentioned division module and the 6th computing module correspond to the step S142 in embodiment 2 To S144, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but it is public to be not limited to the above embodiments 2 institutes The content opened.It should be noted that above-mentioned module can be in such as a group of computer-executable instructions as a part of of device Computer system in execute.

In a kind of optional embodiment, above-mentioned apparatus further includes：6th computing module, for according to each session group pair The first probability and the second probability answered, there are exit priority in each session group of same subscriber for calculating；Sorting module is used Priority is exited in foundation to be ranked up each session group, obtains exiting exiting sequentially for each session group.

Herein it should be noted that above-mentioned 6th computing module and sorting module correspond to the step S152 in embodiment 2 To S154, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but it is public to be not limited to the above embodiments 2 institutes The content opened.It should be noted that above-mentioned module can be in such as a group of computer-executable instructions as a part of of device Computer system in execute.

Embodiment 4

According to the embodiment of the present application, a kind of embodiment of the method exiting group is additionally provided, it is provided in this embodiment to exit group Method can be applied to the management and group in the various immediate communication tools based on internet communication, including it is but unlimited In QQ groups, wechat group, MSN groups, QQ discussion group, the application scenarios such as mailing list.It should be noted that the flow in attached drawing illustrates The step of can be executed in the computer system of such as a group of computer-executable instructions, although also, showing in flow charts Go out logical order, but in some cases, it can be with the steps shown or described are performed in an order that is different from the one herein.

Figure 18 is according to a kind of flow chart of method exiting group of the embodiment of the present application, as shown in figure 18, including as follows Step：

S182 obtains the user behavior characteristics for the multiple users for including in first crowd.

Specifically, in above-mentioned steps, above-mentioned first group can be any one means of communication group, including Instant Messenger Interrogate group (for example, QQ groups, QQ discussion group, wechat group etc.) and non-instant communication group (for example, mailing list etc.)；Due to each group Including multiple users, thus, for a group, user's row in the group of each user in the group can be monitored in real time It is characterized, user behavior characteristics include but not limited to：User participates in the time of session, user exits the operation of the group, shielding The group etc..

S184 obtains at least one user and exits group of probability according to the user behavior characteristics of multiple users.

In getting group after the user behavior characteristics of multiple users, according to the user behavior characteristics of multiple users, obtain Each user in the group is taken to exit the group of probability.

S186, according to probability, it is determined whether group of prompt message is exited in output.

Specifically, in above-mentioned steps, after each user exits the probability of the group in getting group, according to user Exit the probability size of the group, it is determined whether the prompt message of the group is exited to user's output.For example, having when in group After 50% or more user exits the group, it can be determined that the probability that user exits the group is larger, thus can prompt user Exit the group.

From the foregoing, it will be observed that in scheme disclosed in the above embodiments of the present application 4, by detecting certain in communication software in real time or periodically The user behavior characteristics of at least one user in a group, and according to the user behavior characteristics of multiple users in the group, obtain Each user exits the probability of the group in the group, and the probability of the group is then exited according to each user, it is determined whether to The user exports the prompt message for exiting the group, has reached and has predicted to use according to the user behavior characteristics of a large number of users in group The probability of the group is exited at family, to determine whether the purpose for prompting user's execution to move back group operation, is held to realize simplified user Row moves back group operation, improves the technique effect of user experience.

As a result, the scheme of above-described embodiment 4 provided by the present application solve existing instant messaging tools to move back group operation numerous The technical issues of trivial influence user experience effect.

As a kind of optional embodiment, above-mentioned first group can be instant messaging group.Instant messaging group includes but unlimited In：QQ groups, QQ discussion group, wechat group etc..

As in an alternative embodiment, above-mentioned first group can be mailing list.Mailing list includes but not limited to Gmail mailboxes, 163 mailboxes, 126 mailboxes etc..

Embodiment 5

According to the embodiment of the present application, a kind of device for implementing to exit the method for group in above-described embodiment 4 is additionally provided Embodiment, the including but not limited to application scenarios of embodiment 4.Figure 19 is a kind of device exiting group according to the embodiment of the present application Schematic diagram, as shown in figure 19, which includes：First acquisition unit 191, second acquisition unit 193 and determination unit 195.

Wherein, first acquisition unit 191, the user behavior characteristics for obtaining in first group the multiple users for including；

Second acquisition unit 193 exits for according to the user behavior characteristics of multiple users, obtaining at least one user Group of probability；

Determination unit 195, for according to probability, it is determined whether group of prompt message is exited in output.

Herein it should be noted that above-mentioned first acquisition unit 191, second acquisition unit 193 and determination unit 195 correspond to Step S182 to S186 in embodiment 4, above-mentioned module is identical as example and application scenarios that corresponding step is realized, but It is not limited to the above embodiments 4 disclosure of that.It should be noted that above-mentioned module can be such as a part of of device It is executed in the computer system of a group of computer-executable instructions.

From the foregoing, it will be observed that in scheme disclosed in the above embodiments of the present application 5, first acquisition unit 191 is in real time or timing detects In communication software in some group at least one user user behavior characteristics, second acquisition unit 193 is according to more in the group The user behavior characteristics of a user obtain the probability that each user in the group exits the group, and determination unit 195 is according to each User exits the probability of the group, it is determined whether the prompt message that the group is exited to user output has reached according to group The user behavior characteristics of middle a large number of users predict that user exits the probability of the group, and group is moved back to determine whether to prompt user to execute The purpose of operation moves back group operation to realize simplified user's execution, improves the technique effect of user experience.

As a result, the scheme of above-described embodiment 5 provided by the present application solve existing instant messaging tools to move back group operation numerous The technical issues of trivial influence user experience effect.

As a kind of optional embodiment, above-mentioned first group can be instant messaging group or mailing list.

Embodiment 6

Embodiments herein can provide a kind of terminal, which can be in terminal group Any one computer terminal.Optionally, in the present embodiment, above computer terminal can also replace with mobile whole The terminal devices such as end.

Optionally, in the present embodiment, above computer terminal can be located in multiple network equipments of computer network At least one access equipment.

Figure 20 shows a kind of hardware block diagram of terminal.As shown in figure 20, terminal 20 can wrap Include one or more (to use 202a, 202b ... ... in figure, 202n to show) processors 202 (processor 202 may include but Be not limited to the processing unit of Micro-processor MCV or programmable logic device FPGA etc.), memory 204 for storing data, with And the transmitting device 206 for communication function.In addition to this, can also include：(I/O connects for display, input/output interface Mouthful), the port universal serial bus (USB) (can as a port in the port of I/O interfaces by including), network interface, Power supply and/or camera.It will appreciated by the skilled person that structure shown in Figure 20 is only to illustrate, not to above-mentioned The structure of electronic device causes to limit.For example, terminal 20 may also include than shown in Figure 20 more or less groups Part, or with the configuration different from shown in Figure 20.

It is to be noted that said one or multiple processors 202 and/or other data processing circuits lead to herein Can often it be referred to as " data processing circuit ".The data processing circuit all or part of can be presented as software, hardware, firmware Or any other combination.In addition, data processing circuit can be single independent processing module or all or part of be attached to meter In any one in other elements in calculation machine terminal 20.As involved in the embodiment of the present application, data processing electricity Road controls (such as the selection for the variable resistance end path being connect with interface) as a kind of processor.

Processor 202 can call the information and application program of memory storage by transmitting device, to execute following steps Suddenly：Obtain the path chosen in map；According to the traffic information in the path chosen, the dynamic image in path is generated, wherein road The dynamic image of diameter is along the image of the initial position in path to final position dynamic mobile；The dynamic to show paths in map Image.

Memory 204 can be used for storing the software program and module of application software, as automatic in the embodiment of the present application Corresponding program instruction/the data storage device of method of instant communication session group is exited, processor 202 is stored in by operation Software program in reservoir 204 and module realize above-mentioned application to perform various functions application and data processing The method for automatically exiting from instant communication session group of program.Memory 204 may include high speed random access memory, may also include non-easy The property lost memory, such as one or more magnetic storage device, flash memory or other non-volatile solid state memories.At some In example, memory 204 can further comprise the memory remotely located relative to processor 202, these remote memories can To pass through network connection to terminal 20.The example of above-mentioned network includes but not limited to internet, intranet, local Net, mobile radio communication and combinations thereof.

Transmitting device 206 is used to receive via a network or transmission data.Above-mentioned network specific example may include The wireless network that the communication providers of terminal 20 provide.In an example, transmitting device 206 includes that a network is suitable Orchestration (Network Interface Controller, NIC), can be connected with other network equipments by base station so as to Internet is communicated.In an example, transmitting device 206 can be radio frequency (Radio Frequency, RF) module, For wirelessly being communicated with internet.

Display can such as touch-screen type liquid crystal display (LCD), which may make that user can be with The user interface of terminal 20 interacts.

Herein it should be noted that in some optional embodiments, terminal 20 shown in above-mentioned Figure 20 can wrap Include hardware element (including circuit), software element (including the computer code that may be stored on the computer-readable medium) or hardware The combination of both element and software element.It should be pointed out that Figure 20 is only an example of particular embodiment, and it is intended to Show to may be present in the type of the component in above computer terminal 20.

In the present embodiment, above computer terminal 20 can automatically exit from instant communication session group with executing application Method in following steps program code：At least one session group is traversed, the use for each user for including in session group is obtained Family behavioural characteristic；According to the user behavior characteristics for each user for including in session group, obtains each user and exit session group Probability；According to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.

Processor can call the information and application program of memory storage by transmitting device, to execute following step： At least one session group is traversed, the user behavior characteristics for each user for including in session group are obtained；Include according in session group Each user user behavior characteristics, obtain the probability that each user exits session group；According to the probability for exiting session group, really It is fixed whether to export the prompt message for exiting session group.

Optionally, the program code of following steps can also be performed in above-mentioned processor：In real time or timing traverses client circle At least one session group shown in face obtains the user that any one session group is included, wherein session group corresponds to a meeting Talk about window；Detect at least one group's state feature of each user included in session group；If detecting user at least One Population status feature changes, then obtains user's row of the changed user of at least one of session group group's state feature It is characterized.

Optionally, the program code of following steps can also be performed in above-mentioned processor：Reading model file, wherein model File has pre-saved the user behavior characteristics of the session group of at least one client feedback, the user behavior characteristics packet of session group Include the user behavior characteristics of at least one user in session group；According to the user behavior characteristics recorded in model file to session group In include each user user behavior characteristics carry out probability calculation, obtain the probability that each user exits session group.

Optionally, the program code of following steps can also be performed in above-mentioned processor：It obtains and is changed in session group At least one group of state features；It is preceding in the group's state feature being locally stored to update session group variation；Group's state feature is occurred At least one group of state Feature Mappings of the session group of variation are changed to server before session group changes in server Group's state feature；According to updated group's state feature, update in each session group pre-saved in model file at least The user behavior characteristics of one user.

Optionally, the program code of following steps can also be performed in above-mentioned processor：According to the meeting recorded in model file The characteristic value of the user behavior characteristics of group is talked about, the feature of the user behavior characteristics of each user shown in session window is obtained Value；The characteristic value of the user behavior characteristics of each user shown in session window is carried out at calculating using bayesian algorithm Reason, obtains the probability that each user exits each session group.

Optionally, the program code of following steps can also be performed in above-mentioned processor：Application is opened in client end interface At least one session group of software, and collect group's state feature of each session group；If detecting that any one session is mass-sended The event that raw user exits, then preserve the group's state feature for the user that the event of exiting has occurred, and obtains being in and does not exit session group User group's state feature；Wherein, session group number, session group history chat messages number, session group participate in chat number, meeting It is average by average everyone chat messages number, session group in reading number, session group to talk about message in the establishment duration of group, session group Most chat messages numbers, session group are chatted recently by the number of personnel's top set, the nearest chatting time of session group, session group in group Participation number and session group chat messages number in time.

Optionally, the program code of following steps can also be performed in above-mentioned processor：According to session in the open state The event of exiting whether is generated in group to classify to multiple session groups, obtains first kind sample and the second class sample, wherein first Class sample contains in the open state and user and sends out the group's state feature for exiting session group, and the second class sample is in opening State and the group's state feature for not exiting session group；It calculates European between any two group's state feature in first kind sample Distance, and two group's state features for selecting Euclidean distance to be less than the first predetermined threshold carry out the first compression processing, are compressed Third class sample afterwards；Calculate the Euclidean distance between any two group state feature in the second class sample, and select it is European away from The second compression processing is carried out from two group's state features less than the second predetermined threshold, obtains compressed 4th class sample；It will Third class sample and the 4th class sample delivery are to server.

Optionally, the program code of following steps can also be performed in above-mentioned processor：Server is received from multiple visitors The third class sample and the 4th class sample at family end, obtain multiple third class samples and multiple 4th class samples；Calculate multiple thirds Euclidean distance in class sample between any two characteristic selects Euclidean distance to be less than two features of third predetermined threshold One in data is retained, and calculates the Euclidean distance in the 4th class sample between any two characteristic, selects Europe Formula distance is retained less than one in two characteristics of the 4th predetermined threshold；Statistics is retained in multiple third class samples In sample data characteristic value, and count the characteristic value for being retained in sample data in multiple 4th class samples, generate model File.

Optionally, the program code of following steps can also be performed in above-mentioned processor：Multiple third class samples will be retained in In sample data and the sample data that is retained in multiple 4th class samples be respectively mapped to scheduled data area, and mark Label in data area.

Optionally, the program code of following steps can also be performed in above-mentioned processor：It is shown in differentiation client end interface The characteristic value of group's state feature of each session group, obtains the characteristic value of multiple first monoid state features and multiple second monoids The characteristic value of state feature, wherein session group is in the open state and generation is exited and disappears for characterizing for the first monoid state feature Breath, the second monoid state feature are in the open state and do not generate exit message for characterizing session group；Use bayesian algorithm The characteristic value of the characteristic value to multiple first monoid state features and multiple second monoid state features carries out calculation processing respectively, It is in the open state and generate the first probability of exit message to obtain each session group, and each session group is in the open state And the second probability of exit message is not generated.

Optionally, the program code of following steps can also be performed in above-mentioned processor：According to each session group corresponding One probability and the second probability, there are exit priority in each session group of same subscriber for calculating；Foundation exits priority pair Each session group is ranked up, and obtains exiting exiting sequentially for each session group.

It will appreciated by the skilled person that structure shown in Figure 20 is only to illustrate, terminal can also be Smart mobile phone (such as Android phone, iOS mobile phones), tablet computer, applause computer and mobile internet device (Mobile Internet Devices, MID), the terminal devices such as PAD.Figure 20 it does not cause to limit to the structure of above-mentioned electronic device.Example Such as, terminal 20 may also include more than shown in Figure 20 or less component (such as network interface, display device), Or with the configuration different from shown in Figure 20.

One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can To be completed come command terminal device-dependent hardware by program, which can be stored in a computer readable storage medium In, storage medium may include：Flash disk, read-only memory (Read-Only Memory, ROM), random access device (Random Access Memory, RAM), disk or CD etc..

Embodiment 7

Embodiments herein additionally provides a kind of storage medium.Optionally, in the present embodiment, above-mentioned storage medium can For preserving the program code performed by the method for automatically exiting from instant communication session group that above-described embodiment two is provided.

Optionally, in the present embodiment, above-mentioned storage medium can be located in computer network Computer terminal group In any one terminal, or in any one mobile terminal in mobile terminal group.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Time At least one session group is gone through, the user behavior characteristics for each user for including in session group are obtained；Include according in session group The user behavior characteristics of each user, obtain the probability that each user exits session group；According to the probability for exiting session group, determine Whether the prompt message of session group is exited in output.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：It is real When or timing traversal client end interface at least one session group for showing, obtain the user that any one session group is included, Wherein, session group corresponds to a session window；Detect at least one group's state feature of each user included in session group； If detecting that at least one group's state feature of user changes, at least one of session group group's state feature hair is obtained The user behavior characteristics of the user for changing.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：It reads Take model file, wherein model file has pre-saved the user behavior characteristics of the session group of at least one client feedback, meeting The user behavior characteristics of words group include the user behavior characteristics of at least one user in session group；According to what is recorded in model file User behavior characteristics carry out probability calculation to the user behavior characteristics for each user for including in session group, obtain each user and move back Go out the probability of session group.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：It obtains Take at least one group of state features changed in session group；It is special in the group's state being locally stored before update session group variation Sign；At least one group of state Feature Mappings of the session group that group's state feature is changed change server to server Middle session group change before group's state feature；According to updated group's state feature, updates and pre-saved in model file At least one of each session group user user behavior characteristics.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Root According to the characteristic value of the user behavior characteristics of the session group recorded in model file, obtain each user's shown in session window The characteristic value of user behavior characteristics；Using bayesian algorithm to the user behavior characteristics of each user shown in session window Characteristic value carries out calculation processing, obtains the probability that each user exits each session group.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：? At least one session group of application software is opened in client end interface, and collects group's state feature of each session group；If inspection It measures any one session group and the event that user exits occurs, then preserve the group's state feature for the user that the event of exiting has occurred, Obtain group's state feature in the user for not exiting session group；Wherein, session group number, session group history chat messages number, It is average by average every in reading number, session group that session group participates in message in chat number, the establishment duration of session group, session group Most chat messages numbers, session group are chatted by the number of personnel's top set, the nearest of session group in group in people's chat messages number, session group Participation number and session group chat messages number in its time, the nearest chatting time of session group.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Root Classify to multiple session groups according to the event of exiting whether is generated in session group in the open state, obtain first kind sample and Second class sample, wherein first kind sample contains in the open state and user and sends out the group's state feature for exiting session group, Second class sample is group's state feature in the open state and not exiting session group；Calculate any two in first kind sample Euclidean distance between group's state feature, and two group's state features for selecting Euclidean distance to be less than the first predetermined threshold carry out the One compression processing obtains compressed third class sample；Calculate the Europe between any two group state feature in the second class sample Formula distance, and two group's state features for selecting Euclidean distance to be less than the second predetermined threshold carry out the second compression processing, are pressed The 4th class sample after contracting；By third class sample and the 4th class sample delivery to server.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Clothes Business device receives third class sample and the 4th class sample from multiple client, obtains multiple third class samples and the multiple 4th Class sample；It calculates the Euclidean distance between any two characteristic in multiple third class samples, Euclidean distance is selected to be less than the One in two characteristics of three predetermined thresholds is retained, and calculate the 4th class sample in any two characteristic it Between Euclidean distance, select Euclidean distance be less than the 4th predetermined threshold two characteristics in one retained；Statistics It is retained in the characteristic value of the sample data in multiple third class samples, and counts the sample number being retained in multiple 4th class samples According to characteristic value, generate model file.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：It will The sample data being retained in multiple third class samples and the sample data being retained in multiple 4th class samples are respectively mapped to Scheduled data area, and the label within the scope of flag data.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Area Divide the characteristic value of group's state feature of each session group shown in client end interface, obtains multiple first monoid state features The characteristic value of characteristic value and multiple second monoid state features, wherein the first monoid state feature is in for characterizing session group Opening state and exit message is generated, the second monoid state feature is in the open state and do not generate and exit for characterizing session group Message；Use the bayesian algorithm characteristic value to multiple first monoid state features and multiple second monoid state features respectively Characteristic value carries out calculation processing, and it is in the open state and generate the first probability of exit message and every to obtain each session group A session group is in the open state and does not generate the second probability of exit message.

Optionally, in the present embodiment, storage medium is arranged to store the program code for executing following steps：Root According to corresponding first probability of each session group and the second probability, there are exiting in each session group of same subscriber is preferential for calculating Grade；Each session group is ranked up according to priority is exited, obtains exiting exiting sequentially for each session group.

Embodiment 8

Embodiments herein additionally provides a kind of system, including：Processor and memory, wherein memory and processing Connection, for providing the instruction for handling following processing step for processor：

Step 402, at least one session group is traversed, the user behavior characteristics for each user for including in session group are obtained；

Step 404, it according to the user behavior characteristics for each user for including in session group, obtains each user and exits session The probability of group；

Step 406, according to the probability for exiting session group, it is determined whether the prompt message of session group is exited in output.It is optional Ground, in the present embodiment, what above-mentioned storage medium can be used to save that above-described embodiment two provided automatically exits from instant messaging Program code performed by the method for session group.

It should be noted that the scene that above-mentioned steps S402 to S406 is applied includes but not limited to the embodiment of the present application two Disclosure of that.

Above-mentioned the embodiment of the present application serial number is for illustration only, can not represent the quality of embodiment.

In above-described embodiment of the application, all emphasizes particularly on different fields to the description of each embodiment, do not have in some embodiment The part of detailed description may refer to the associated description of other embodiment.

In several embodiments provided herein, it should be understood that disclosed technology contents can pass through others Mode is realized.Wherein, the apparatus embodiments described above are merely exemplary, for example, the unit division, only A kind of division of logic function, formula that in actual implementation, there may be another division manner, such as multiple units or component can combine or Person is desirably integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual Between coupling, direct-coupling or communication connection can be INDIRECT COUPLING or communication link by some interfaces, unit or module It connects, can be electrical or other forms.

The unit illustrated as separating component may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, you can be located at a place, or may be distributed over multiple In network element.Some or all of unit therein can be selected according to the actual needs to realize the mesh of this embodiment scheme 's.

In addition, each functional unit in each embodiment of the application can be integrated in a processing unit, it can also It is that each unit physically exists alone, it can also be during two or more units be integrated in one unit.Above-mentioned integrated list The form that hardware had both may be used in member is realized, can also be realized in the form of SFU software functional unit.

If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product When, it can be stored in a computer read/write memory medium.Based on this understanding, the technical solution of the application is substantially The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words It embodies, which is stored in a storage medium, including some instructions are used so that a computer Equipment (can be personal computer, server or network equipment etc.) execute each embodiment the method for the application whole or Part steps.And storage medium above-mentioned includes：USB flash disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic disc or CD etc. are various can to store program code Medium.

The above is only the preferred embodiment of the application, it is noted that for the ordinary skill people of the art For member, under the premise of not departing from the application principle, several improvements and modifications can also be made, these improvements and modifications are also answered It is considered as the protection domain of the application.

Claims

1. a kind of system automatically exiting from instant communication session group, which is characterized in that including：

Server, for providing model file, wherein the model file has pre-saved the meeting of at least one client feedback The user behavior characteristics of group are talked about, the user behavior characteristics of the session group include the use of at least one of session group user Family behavioural characteristic；

Client, and the server communication obtain in the session group for traversing at least one session group and each of include The user behavior characteristics of user, and according to the user behavior characteristics for each user for including in the session group, obtain each use The probability of the session group is exited at family, according to the probability for exiting the session group, it is determined whether the session is exited in output The prompt message of group.

2. a kind of method automatically exiting from instant communication session group, which is characterized in that including：

At least one session group is traversed, the user behavior characteristics for each user for including are obtained in the session group；

According to the user behavior characteristics for each user for including in the session group, obtains each user and exit the session group Probability；

According to the probability for exiting the session group, it is determined whether the prompt message of the session group is exited in output.

3. according to the method described in claim 2, it is characterized in that, traverse at least one session group, obtain in the session group Including each user user behavior characteristics, including：

At least one session group shown in real time or in timing traversal client end interface, obtaining any one session group is included User, wherein the session group corresponds to a session window；

Detect at least one group's state feature of each user included in the session group；

If detecting that at least one group's state feature of user changes, at least one of described session group group's shape is obtained The user behavior characteristics of the changed user of state feature.

4. according to the method described in claim 3, it is characterized in that, according to the user for each user for including in the session group Behavioural characteristic obtains the probability that each user exits the session group, including：

Reading model file, wherein the model file has pre-saved the user of the session group of at least one client feedback Behavioural characteristic, the user behavior characteristics of the session group include the user behavior characteristics of at least one user in the session group；

User's row according to the user behavior characteristics recorded in the model file to each user for including in the session group It is characterized carry out probability calculation, obtains the probability that each user exits the session group.

5. according to the method described in claim 4, it is characterized in that, if detecting at least one group's state feature hair of user Changing then obtains the user behavior characteristics of the changed user of at least one of described session group group's state feature, including：

Obtain at least one group of state features changed in the session group；

It is preceding in the group's state feature being locally stored to update the session group variation；

At least one group of state Feature Mappings of the session group that group's state feature is changed change institute to server State group's state feature before session group described in server changes；

According to updated group's state feature, at least one of each session group pre-saved in the model file is updated The user behavior characteristics of user.

6. according to the method described in claim 5, it is characterized in that, according to the user behavior characteristics recorded in the model file Value carries out probability calculation to the user behavior characteristics of each user for including in the session group, obtain each user exit it is described The probability of session group, including：

According to the characteristic value of the user behavior characteristics of the session group recorded in the model file, obtains in the session window and show The characteristic value of the user behavior characteristics of each user shown；

The characteristic value of the user behavior characteristics of each user shown in the session window is counted using bayesian algorithm Calculation is handled, and obtains the probability that each user exits each session group.

7. according to the method described in claim 6, which is characterized in that detecting each user included in the session group At least one group's state feature before, the method further includes：

At least one session group of application software is opened in the client end interface, and collects group's shape of each session group State feature；

If detecting that the event that user exits occurs for any one session group, the group for the user that the event of exiting has occurred is preserved State feature obtains group's state feature in the user for not exiting the session group；

8. the method according to the description of claim 7 is characterized in that group's shape of the user for the event of exiting has occurred described in preservation State feature, after obtaining group's state feature of each session group in the open state, the method further includes：

It is transmitted to the server after group's state feature of each session group in the open state is carried out compression processing, The step includes：

At least one session group is divided according to the event of exiting whether is generated in the session group in the open state Class obtains first kind sample and the second class sample, wherein the first kind sample is contained in the opening state and user The group's state feature for exiting the session group is sent out, the second class sample is not exit in the opening state and described Group's state feature of session group；

The Euclidean distance between any two group state feature in the first kind sample is calculated, and Euclidean distance is selected to be less than the Two group's state features of one predetermined threshold carry out the first compression processing, obtain compressed third class sample；

The Euclidean distance between any two group state feature in the second class sample is calculated, and Euclidean distance is selected to be less than the Two group's state features of two predetermined thresholds carry out the second compression processing, obtain compressed 4th class sample；

By the third class sample and the 4th class sample delivery to the server.

9. according to the method described in claim 8, it is characterized in that, being passed by the third class sample and the 4th class sample It transports to after the server, the method further includes：

The server receives third class sample and the 4th class sample from multiple client, obtains multiple third class samples With multiple 4th class samples；

It calculates the Euclidean distance between any two characteristic in the multiple third class sample, Euclidean distance is selected to be less than the One in two characteristics of three predetermined thresholds is retained, and calculates any two characteristic in the 4th class sample Euclidean distance between, select Euclidean distance to be less than in two characteristics of the 4th predetermined threshold one are retained；

Statistics is retained in the characteristic value of the sample data in the multiple third class sample, and counts and be retained in the multiple 4th The characteristic value of sample data in class sample generates the model file.

10. according to the method described in claim 9, it is characterized in that, the characteristic value is returned for characterizing the multiple client The frequency of occurrences, the occurrence number of the group's state feature for each session group returned, wherein be retained in the multiple third class in statistics The characteristic value of sample data in sample, and the characteristic value for being retained in the sample data in the multiple 4th class sample is counted, Before generating the model file, the method further includes：

By the sample data being retained in the multiple third class sample and the sample being retained in the multiple 4th class sample Data are respectively mapped to scheduled data area, and the label within the scope of flag data.

11. the method according to any one of claim 6 to 10, which is characterized in that using bayesian algorithm to described The characteristic value of group's state feature of each session group shown in client end interface carries out calculation processing, obtains each user and exits The probability of each session group, including：

The characteristic value of group's state feature of each session group shown in the client end interface is distinguished, multiple first monoids are obtained The characteristic value of the characteristic value of state feature and multiple second monoid state features, wherein the first monoid state feature is used for It is in the open state and generate exit message to characterize each session group, the second monoid state feature is described for characterizing Each session group is in the open state and does not generate exit message；

Use the bayesian algorithm characteristic value to the multiple first monoid state feature and the multiple second class respectively The characteristic value of group's state feature carries out calculation processing, and it is in the open state and generate exit message to obtain each session group First probability and each session group are in the open state and do not generate the second probability of exit message.

12. according to the method for claim 11, which is characterized in that exit the general of each session group according to each user Rate is ranked up all session groups, obtains the sequence that exits for exiting each session group, including：

According to corresponding first probability of each session group and the second probability, calculate that there are in each session group of same subscriber Exit priority；

It exits priority according to described in each session group is ranked up, obtain exiting exiting for each session group suitable Sequence.

13. a kind of device automatically exiting from instant communication session group, which is characterized in that including：

First acquisition module obtains in the session group user for each user for including for traversing at least one session group Behavioural characteristic；

Second acquisition module obtains each use for the user behavior characteristics according to each user for including in the session group The probability of the session group is exited at family；

Determining module, the probability for exiting the session group according to, it is determined whether carrying for the session group is exited in output Show information.

14. a kind of method exiting group, which is characterized in that including：

Obtain the user behavior characteristics for the multiple users for including in first crowd；

According to the user behavior characteristics of the multiple user, obtains at least one user and exit group of probability；

According to the probability, it is determined whether the group of prompt message is exited in output.

15. according to the method for claim 14, which is characterized in that described first group is instant messaging group.

16. according to the method for claim 14, which is characterized in that described first group is mailing list.

17. a kind of device exiting group, which is characterized in that including：

First acquisition unit, the user behavior characteristics for obtaining in first group the multiple users for including；

Second acquisition unit obtains at least one user and exits first for the user behavior characteristics according to the multiple user The probability of group；

Determination unit, for according to the probability, it is determined whether the group of prompt message is exited in output.

18. a kind of storage medium, which is characterized in that the storage medium includes the program of storage, wherein run in described program When control the storage medium where equipment perform claim require to automatically exit from instant messaging meeting described in any one of 2 to 12 The method for talking about group.

19. a kind of processor, which is characterized in that the processor is for running program, wherein right of execution when described program is run Profit requires the method for automatically exiting from instant communication session group described in any one of 2 to 12.

20. a kind of system, which is characterized in that including：

Processor；And

Memory is connected to the processor, for providing the instruction for handling following processing step for the processor：

Step S402 traverses at least one session group, obtains in the session group user behavior characteristics for each user for including；

Step S404 obtains each user and exits institute according to the user behavior characteristics for each user for including in the session group State the probability of session group；

Step S406, according to the probability for exiting the session group, it is determined whether the prompt letter of the session group is exited in output Breath.