CN106844724A - It is a kind of that the method and apparatus applied to recommend application have been installed based on user - Google Patents

It is a kind of that the method and apparatus applied to recommend application have been installed based on user Download PDF

Info

Publication number
CN106844724A
CN106844724A CN201710072980.XA CN201710072980A CN106844724A CN 106844724 A CN106844724 A CN 106844724A CN 201710072980 A CN201710072980 A CN 201710072980A CN 106844724 A CN106844724 A CN 106844724A
Authority
CN
China
Prior art keywords
application
user
similarity
preference
preset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201710072980.XA
Other languages
Chinese (zh)
Other versions
CN106844724B (en
Inventor
潘岸腾
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alibaba China Co Ltd
Original Assignee
Guangzhou Youshi Network Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Youshi Network Technology Co Ltd filed Critical Guangzhou Youshi Network Technology Co Ltd
Priority to CN201710072980.XA priority Critical patent/CN106844724B/en
Publication of CN106844724A publication Critical patent/CN106844724A/en
Application granted granted Critical
Publication of CN106844724B publication Critical patent/CN106844724B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Stored Programmes (AREA)

Abstract

The invention provides a kind of method and apparatus installed based on user and applied to recommend application.Methods described includes：Determine the similarity list between applying two-by-two in preset application library；Based on mounted one or more applications of user and the similarity list, preference of the user to each application in preset application library is determined；A number of respective application is chosen from preset application library based on the preference angle value descending order to be applied as recommendation.

Description

It is a kind of that the method and apparatus applied to recommend application have been installed based on user
Technical field
The present invention relates to technical field of information processing, application is installed based on user in particular to one kind to recommend to answer Method and apparatus.
Background technology
With the fast development of Internet technology and intelligent mobile terminal technology, the work(much realized on computer terminals (can for example do shopping, read) can also realize on intelligent mobile terminal, for example, use smart mobile phone or panel computer etc..Separately Outward, the realization of these functions needs to install corresponding application program on intelligent mobile terminal.For example, shopping online is, it is necessary to pacify Dress such as Taobao's client, listens music to need to install music player client etc..Thus, many software companys provide application Shop or application market, such as pea pods or PP assistant etc..User can be opened using shop or application market, so as to Various application programs required for enough fast searchs and download, including audio-visual broadcast message class, system tool class, the social class of communication, net Upper shopping class, reading class etc., certainly can be with amusement and recreation class application program such as download games (APP).
Application shop or application market in, in order to constantly lifted user using application shop or application market it is good Good experience sense, current developer develops the function that many convenient users are used, and one of them is recommendation function, i.e., pushed away to user Some applications are recommended, to help user to find more applications interested, for example：" guessing that you like ", " everybody also downloads " etc..At present Recommendation application method it is basic recommended with the popular ratings applied, for example recommend download to come above Using, or recommend the application in popular ranking list.But because different users possesses different interest, according to prior art It is interested that the application of recommendation is not necessarily user, it is impossible to meets the demand of different user, causes the experience sense of user not good.
The content of the invention
It is an object of the invention to provide a kind of method and apparatus installed based on user and applied to recommend application, to change Kind above mentioned problem.
A kind of method installed based on user and applied to recommend application is the embodiment of the invention provides, it includes：
Determine the similarity list between applying two-by-two in preset application library；
Based on mounted one or more applications of user and the similarity list, determine the user to preset application library In each application preference；
A number of respective application conduct is chosen from preset application library based on the preference angle value descending order Recommend application.
It is determined that the user in preset application library each application preference and based on the preference angle value from greatly to Small order from chosen in preset application library a number of respective application as recommend application the step of in, from preset application library Multiple applications are selected as pre- recommendation application, is set up comprising the plurality of pre- Candidate Set for recommending to apply, thereby determine that the user The preference for recommending to apply pre- to each in the Candidate Set, based on the preference angle value descending order from the candidate A number of respective application is chosen in collection to be applied as recommendation.
Preferably, the pre- recommendation application is the label same label with one or more applications mounted with user Application.
The embodiment of the present invention additionally provides a kind of device installed based on user and applied to recommend to apply, and it includes：
The similarity list determining unit of application, for determining the similarity row between the application two-by-two in preset application library Table；
User to apply preference determining unit, for based on user it is mounted one or more application and the phase Like degree list, preference of the user to each application in preset application library is determined；
Recommendation unit, for choosing a number of from preset application library based on the preference angle value descending order Respective application is applied as recommendation.
Preferably, the Similarity value between applying two-by-two in preset application library is calculated, between these are applied two-by-two Similarity value is made list, it is described apply two-by-two between similarity computational methods it is as follows：
Wherein：K=1,2 ..., n j=1,2 ..., n
Sk,lRepresent the similarity between application k and application l；
N represents the number of applications in preset application library；
M represents the number of labels in preset tag set；
tk,jRepresent using whether k has label j, promising 1, inaction 0；
tl,jRepresent using whether l has label j, promising 1, inaction 0；
rjThe resolution ratio of label j is represented, wherein：
N represents the number of applications in preset application library；
ti,jRepresent using whether i has label j, promising 1, inaction 0.
The preference that the user is applied to certain for the mounted multiple applications of user respectively with it is described certain apply it Between similarity cumulative sum, computing formula is as follows：
Wherein：PlRepresent preference of the user to the application l in preset application library；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and preset application library.
The embodiment of the present invention additionally provides a kind of device installed based on user and applied to recommend to apply, and it includes：
The similarity list determining unit of application, for determining the similarity row between the application two-by-two in preset application library Table；
The pre- Candidate Set for recommending application sets up unit, for selecting multiple applications from preset application library as pre- recommendation Using foundation includes the plurality of pre- Candidate Set for recommending to apply；
User to apply preference determining unit, for based on user it is mounted one or more application and the phase Like degree list, the user preference for recommending to apply pre- to each in the Candidate Set is determined；
Recommendation unit, for choosing a number of from the Candidate Set based on the preference angle value descending order Respective application is applied as recommendation.
Preferably, the pre- recommendation application is the label same label with one or more applications mounted with user Application.
It is of the invention that the method and apparatus applied to recommend application have been installed based on user, initially set up preset application Similarity list between applying two-by-two in storehouse；Then according to mounted one or more applications of user and the similarity List come determine the user in preset application library each application preference；Suitable from big to small according to the preference angle value Sequence chooses a number of respective application from preset application library as recommending to apply, it is achieved thereby that the interest love according to user The purpose of personalized recommendation application is carried out well, substantially increases Consumer's Experience.
Brief description of the drawings
Fig. 1 be first embodiment of the invention installed based on user using come recommend application method flow chart；
Fig. 2 be second embodiment of the invention installed based on user using come recommend application method flow chart；
Fig. 3 be third embodiment of the invention installed based on user using come recommend application device schematic frame Figure；
Fig. 4 be fourth embodiment of the invention installed based on user using come recommend application device schematic frame Figure.
Specific embodiment
Below in conjunction with the embodiment of the present invention and accompanying drawing, the technical scheme in the embodiment of the present invention is carried out clear, complete Ground description, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Generally herein The component of the embodiment of the present invention described and illustrated in place's accompanying drawing can be arranged and designed with a variety of configurations.Therefore, The detailed description of the embodiments of the invention to providing in the accompanying drawings is not intended to limit the model of claimed invention below Enclose, but be merely representative of selected embodiment of the invention.Based on embodiments of the invention, those skilled in the art are not making The every other embodiment obtained on the premise of creative work, belongs to the scope of protection of the invention.
One of the reason for Consumer's Experience sense being generally noted above is not good is that different users possesses different interest, and existing The suggested design of technology is only the application for recommending download to come above, but download highest application is not necessarily people What people liked.By taking game as an example, it is assumed that " fishing intelligent APP " is download highest, but user A does not like object for appreciation " fishing reaches The game of people ", but like playing FTG, so it is that his interest cannot be excited to go a little to recommend " fishing intelligent " to user A Hit download；Again for example, user B has downloaded " fishing intelligent " under the recommendation of friend, but object for appreciation is not liked, unloaded.But according to The recommendation method of prior art, when user B click to enter using shop or application market when still can to its recommendation " fishing reaches People ", this just brings bad Consumer's Experience.
The inventor of the technical program has taken into full account the interest and hobby of user, proposes a kind of new personalized recommendation side Method, can recommend different application according to the hobby of different user difference, and so as to realize personalized recommendation, this can be carried significantly Rise the experience sense of user.
Generally it can be thought that, installed on the intelligent terminal such as such as smart mobile phone or panel computer that user uses or computer Various applications, such as game class, leisure, office class, be user application interested, if it is possible to find a kind of method Application can be installed based on user to recommend application, it is possible to realize the purpose of described personalized recommendation.
Fig. 1 be first embodiment of the invention installed based on user using come recommend application method flow chart.Such as Shown in Fig. 1, the method that application has been installed based on user to recommend application of the invention is comprised the following steps：
S1：Determine the similarity list between applying two-by-two in preset application library.
Development and application market or application shop when generally all can preset application library, from application market or application shop download Application message all in the preset application library is stored in.The method of the present invention is to calculate two-by-two should in preset application library first Similarity value between, the Similarity value between these are applied two-by-two is made list.The similarity list can include：Two No. ID of individual application, the Similarity value between the two applications.For example shown in following table：
First applies ID Second applies ID Similarity
00001 00002 0.21
00001 00003 0.35
00001 00004 0.26
… … …
00002 00003 0.13
00002 00004 0.29
00002 00005 0.38
… … …
It is determined that the method for the similarity between applying two-by-two has a lot, very simple method such as classification, by similar two The similarity of individual application is set to 1, and the similarity of inhomogeneous two applications is set to 0.In addition, using in shop or application market The various third party applications (referred to as using) for providing generally all have label, and the effect of label is the various application programs of mark Classification or content, be easy to user to search.At present, each application can include at least 1 in application market or application shop Individual application label, so can determine its similarity between two according to whether application two-by-two has same label.Furthermore, The value of similarity can also be determined according to the quantity with same label, for example, there can be 1 the two of same label to answer with setting tool Similarity between is 1, and it is 2 that setting tool has the similarity between 2 the two of same label applications.Certainly, illustrate here Similarity value is 1 or 2 etc., is merely for illustrative purposes, can be digital using other in practice, for example (0,1] between Percent value.
The third-party application provided using shop or application market all has one or more labels, and these labels come from The preset tag set at development and application shop or application market, this is known to the skilled person general knowledge, Here these routine techniques are not done with excessive introduction.
Certainly, the method for the example above is simplest method, it is also possible to use other method.But it is excellent in the present invention Choosing provides one kind and more preferably determines method, and thus obtained Similarity value can be more showed between applying two-by-two in preset application library Similitude.
It is described apply two-by-two between similarity computational methods it is as follows：
Wherein：K=1,2 ..., n j=1,2 ..., n
Sk,lRepresent the similarity between application k and application l；
N represents the number of applications in preset application library；
M represents the number of labels in preset tag set；
tk,jRepresent using whether k has label j, promising 1, inaction 0；
tl,jRepresent using whether l has label j, promising 1, inaction 0；
rjThe resolution ratio of label j is represented, wherein：
N represents the number of applications in preset application library；
ti,jRepresent using whether i has label j, promising 1, inaction 0.
Used here as the resolution ratio r of label jjThe fine degree of content is divided by label j to reflect, resolution value is bigger Represent that division content is finer.
For example：There are two labels " flashlight ", " small tools " in preset tag set.The application for having " flashlight " has 10 sections, the function of these applications is all mobile phone illumination.The application for having " small tool " label has 100 sections, wherein except comprising institute Outside thering is " flashlight " to apply, some applications on the small tool class such as weather lookup, alarm clock, calculator are further comprises.So Label " flashlight " is more finer than label " small tool " on content is divided in this example embodiment, by the label of " flashlight " Related application more can be exactly matched, it can be considered that the resolution ratio of " flashlight " is higher.Can be with quantization resolution Weighed by the inverse for counting the quantity of the application that a label is covered.In this example embodiment, the resolution ratio of " flashlight " is 0.1, the rate respectively of " small tool " is 0.01, it is believed that the resolution ratio of " flashlight " is 10 times of the resolution ratio of " small tool ", It is finer that it divides content.
There is the resolution ratio of same label with using l using k in being applied two-by-two here by superposition, then be standardized place Reason, Similarity value is compressed to (0,1] between.
S2：Based on mounted one or more applications of user and the similarity list, determine the user to preset application The preference of each application in storehouse.
It refers to have pacified in the terminal that user uses when recommending and applying to user that user described here has installed and applied The third-party application being filled with.
Want to be found out from preset application library and be adapted to the multiple applications to user's recommendation, it is necessary to one can determine the user To the method for the preference of each application in preset application library, the size according to preference recommends application to choose.
Here the method for using for：The mounted multiple applications of user are tired with the similarity between certain application respectively In addition and as the user to it is described certain apply preference value.That is, from the similarity row set up in previous step Some in preset application library is found out in table using the similarity between one or more applications mounted with user respectively Value, by the cumulative summation of the plurality of Similarity value, institute's value is preference angle value of the user to the application.Similarly, can obtain Preference of the user to each application in preset application library.Determine that the computing formula of the preference is as follows：
Wherein：PlRepresent preference of the user to the application l in preset application library；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and preset application library.
Due to refer here to using user installed application and preset application library in application between similarity, so this In the user installed using include user by application shop or application market download install application and can apply The application found in shop or application market.
S3：A number of respective application is chosen from preset application library based on the preference angle value descending order to make To recommend to apply.
According to resulting user in preset application library each apply preference, by preference angle value from big to small Order chooses a number of respective application from preset application library, such as down chooses certain amount since preference maximum Respective application as recommend apply, show user.The certain amount can for example select 5 with unrestricted choice in practice Individual, 10,20,50 or its value etc..
In the above embodiments, using the side for determining the preference that the user applies to each in preset application library Number of applications in formula, but preset application library is a lot, and amount of calculation can be caused very big.A preferred embodiment is described below, in advance Part application is selected from preset application library as pre- recommendation application, is set up comprising the plurality of pre- candidate for recommending to apply Collection, thereby determines that the user preference for recommending to apply pre- to each in the Candidate Set, can reduce amount of calculation, and raising is looked into Look for the speed for recommending to apply.
Fig. 2 be second embodiment of the invention installed based on user using come recommend application method flow chart.Such as Shown in Fig. 2, the method that application has been installed based on user to recommend application of the invention is comprised the following steps：
S1：Determine the similarity list between applying two-by-two in preset application library.
Corresponding steps S1 with first embodiment is identical, and description is not repeated.
S2：Multiple applications are selected from preset application library and recommends application as pre-, setting up should comprising the plurality of pre- recommendation Candidate Set.
Multiple applications are selected from preset application library can use various ways as the pre- choosing method for recommending application Or various ways and deposit.Such as one of choosing method：The download of the application that statistics application shop or application market are provided, choosing Remove carrying capacity multiple applications in the top and recommend application as pre-, or it is many from the classification identical that application has been installed with user In individual classification application, choose the multiple applications in the top of download in each classification and recommend application as pre-；Choosing method it Two：The quality score of the application that statistics application shop or application market are provided, chooses scoring multiple application conducts high and pushes away in advance Application is recommended, or from the classification identical multiple classification application that application has been installed with user, to be chosen and score high in each classification Multiple applications recommend application as pre-；The three of choosing method：The application that statistics application shop or application market are provided turn Rate, that is, it is conversion ratio to click on an application and download its number of users with the ratio of the number of users for clicking on the application, chooses and turns Rate multiple application conducts in the top are pre- to recommend application, or divides from the classification identical that application has been installed with user is multiple In class application, choose the multiple applications in the top of conversion ratio in each classification and recommend application as pre-；The four of choosing method：Choosing The application for taking the label same labels with one or more applications mounted with user recommends application as pre-, if that is, with Mounted 1 application in family has 3 labels, then choose the application with 1 or 2 or 3 label in 3 labels and make Recommend application for pre-.In addition, it is possible to use other method is chosen, and can also carry out any combinations to above-mentioned 4 kinds of methods Recommend application as pre- to select multiple applications from preset application library.
Set up comprising the plurality of pre- Candidate Set for recommending to apply.For the multiple application for selecting, they can be recorded It is thousands of or tens of thousands of if the pre- recommendation number of applications chosen is many in a list, it is also possible to which that individually setting up a candidate should With storehouse, the selected pre- recommendation application of the inside record.
S3：Based on mounted one or more applications of user and the similarity list, determine the user to the candidate Each pre- preference for recommending to apply in collection.
Corresponding steps S2 with first embodiment is identical.That is, the time is found out from the similarity list set up Some in selected works, will be the plurality of similar using the Similarity value between one or more applications mounted with user respectively The cumulative summation of angle value, institute's value is preference angle value of the user to the application.Similarly, user can be obtained to the Candidate Set In each application preference.Determine that the computing formula of the preference is as follows：
Wherein：PlRepresent preference of the user to the application l in Candidate Set；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and the Candidate Set.
Due to refer here to using user installed application and preset application library in application between similarity, so this In the user installed using include user by application shop or application market download install application and can apply The application found in shop or application market.
S4：A number of respective application is chosen from the Candidate Set based on the preference angle value descending order to make To recommend to apply.
Corresponding steps S3 with first embodiment is identical, and description is not repeated.
Compared with first embodiment, difference part is pre- from preset application library based on certain condition to the second embodiment First select multiple applications and recommend application as pre-, the applications of of poor quality, unmanned concern can be given up, it is to avoid follow-up calculating is used Family reduces amount of calculation to the preference of these ropy applications, improves the speed searched and recommend application.
It is of the invention that the method applied to recommend application has been installed based on user, can determine that the user is answered preset The preference applied with each in storehouse；One fixed number is chosen from preset application library according to the preference angle value descending order The respective application of amount realizes the purpose that personalized recommendation application is carried out according to the hobby of user, greatly as recommending to apply Improve Consumer's Experience greatly.
Fig. 3 be third embodiment of the invention installed based on user using come recommend application device schematic frame Figure.As shown in figure 3, the device for having been installed application based on user to recommend application of the invention is included：
The similarity list determining unit of application, for determining the similarity row between the application two-by-two in preset application library Table；
User to apply preference determining unit, for based on user it is mounted one or more application and the phase Like degree list, preference of the user to each application in preset application library is determined；
Recommendation unit, for choosing a number of from preset application library based on the preference angle value descending order Respective application is applied as recommendation.
Preferably, the similarity list determining unit of the application is used to calculating between applying two-by-two in preset application library Similarity value, the Similarity value between these are applied two-by-two is made list, it is described apply two-by-two between similarity meter Calculation method is as follows：
Wherein：K=1,2 ..., n j=1,2 ..., n
Sk,lRepresent the similarity between application k and application l；
N represents the number of applications in preset application library；
M represents the number of labels in preset tag set；
tk,jRepresent using whether k has label j, promising 1, inaction 0；
tl,jRepresent using whether l has label j, promising 1, inaction 0；
rjThe resolution ratio of label j is represented, wherein：
N represents the number of applications in preset application library；
ti,jRepresent using whether i has label j, promising 1, inaction 0.
Preferably, the user preference determining unit applied is used for by the mounted multiple applications of user respectively with The cumulative sum of the similarity between the application is as the user to the preference of the application, and computing formula is as follows：
Wherein：PlRepresent preference of the user to the application l in preset application library；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and preset application library.
It is apparent to those skilled in the art that, for convenience and simplicity of description, with reference to 3rd embodiment The specific work process of the device of description, may be referred to the corresponding process in aforementioned first embodiment, and description is not repeated herein.
Fig. 4 be fourth embodiment of the invention installed based on user using come recommend application device schematic frame Figure.As shown in figure 4, the device for having been installed application based on user to recommend application of the invention is included：
The similarity list determining unit of application, for determining the similarity row between the application two-by-two in preset application library Table；
The pre- Candidate Set for recommending application sets up unit, for selecting multiple applications from preset application library as pre- recommendation Using foundation includes the plurality of pre- Candidate Set for recommending to apply；
User to apply preference determining unit, for based on user it is mounted one or more application and the phase Like degree list, the user preference for recommending to apply pre- to each in the Candidate Set is determined；
Recommendation unit, for choosing a number of from the Candidate Set based on the preference angle value descending order Respective application is applied as recommendation.
Preferably, the work of the similarity list determining unit of application is identical with what 3rd embodiment was described.
Preferably, the course of work that the pre- Candidate Set for recommending to apply sets up unit may be referred in aforementioned second embodiment Corresponding process, is not repeated description herein, and the pre- recommendation application preferably has one or more applications mounted with user Label same label application.
Preferably, during the course of work of preference determining unit of the user to applying may be referred to aforementioned second embodiment Corresponding process.That is, found out from the similarity list set up in the Candidate Set some application respectively with user Similarity value between mounted one or more applications, by the cumulative summation of the plurality of Similarity value, institute's value is the user To the preference angle value of the application.Similarly, preference of the user to each application in the Candidate Set can be obtained.Determine institute The computing formula for stating preference is as follows：
Wherein：PlRepresent preference of the user to the application l in Candidate Set；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and the Candidate Set.
Due to refer here to using user installed application and preset application library in application between similarity, so this In the user installed using include user by application shop or application market download install application and can apply The application found in shop or application market.
It is apparent to those skilled in the art that, for convenience and simplicity of description, with reference to fourth embodiment The specific work process of the device of description, may be referred to the corresponding process in aforementioned second embodiment, and description is not repeated herein.
It is of the invention that the device applied to recommend to apply has been installed based on user, can determine that the user is answered preset The preference applied with each in storehouse；One fixed number is chosen from preset application library according to the preference angle value descending order The respective application of amount realizes the purpose that personalized recommendation application is carried out according to the hobby of user, greatly as recommending to apply Improve Consumer's Experience greatly.
What the embodiment of the present invention was provided has installed application based on user to recommend the computer program product of the method for application Product, including the computer-readable recording medium of program code is stored, before the instruction that described program code includes can be used to perform Method described in the embodiment of the method for face, implements and can be found in embodiment of the method, will not be repeated here.
If the function is to realize in the form of SFU software functional unit and as independent production marketing or when using, can be with Storage is in a computer read/write memory medium.Based on such understanding, technical scheme is substantially in other words The part contributed to prior art or the part of the technical scheme can be embodied in the form of software product, the meter Calculation machine software product is stored in a storage medium, including some instructions are used to so that a computer equipment (can be individual People's computer, panel computer, smart mobile phone, server, or network equipment etc.) perform each embodiment methods described of the invention All or part of step.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM), arbitrary access are deposited Reservoir (RAM), magnetic disc or CD etc. are various can be with the medium of store program codes.
The above, specific embodiment only of the invention, but protection scope of the present invention is not limited thereto, and it is any Those familiar with the art the invention discloses technical scope in, change or replacement can be readily occurred in, should all contain Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.

Claims (10)

1. a kind of that the method applied to recommend application has been installed based on user, it includes：
Determine the similarity list between applying two-by-two in preset application library；
Based on mounted one or more applications of user and the similarity list, determine the user in preset application library The preference of each application；
A number of respective application is chosen from preset application library as recommendation based on the preference angle value descending order Using.
2. method according to claim 1, it is characterised in that it is determined that the user is to each application in preset application library Preference and a number of respective application chosen from preset application library based on the preference angle value descending order make In the step of to recommend application, multiple applications are selected from preset application library and recommends application as pre-, set up comprising the plurality of The pre- Candidate Set for recommending to apply, thereby determines that the user preference for recommending to apply pre- to each in the Candidate Set, is based on The preference angle value descending order is chosen a number of respective application from the Candidate Set and is applied as recommendation.
3. method according to claim 1, it is characterised in that it is determined that phase between applying two-by-two in preset application library In like the step of degree list, the Similarity value between applying two-by-two in preset application library is calculated, between these are applied two-by-two Similarity value be made list, it is described apply two-by-two between similarity computational methods it is as follows：
S k , l = Σ j = 1 m ( r j * t k , j * t l , j ) Σ j = 1 m ( r j * t k , j ) k ≠ l
Wherein：K=1,2 ..., n j=1,2 ..., n
Sk,lRepresent the similarity between application k and application l；
N represents the number of applications in preset application library；
M represents the number of labels in preset tag set；
tk,jRepresent using whether k has label j, promising 1, inaction 0；
tl,jRepresent using whether l has label j, promising 1, inaction 0；
rjThe resolution ratio of label j is represented, wherein：
r j = 1 Σ i = 1 n t i , j
N represents the number of applications in preset application library；
ti,jRepresent using whether i has label j, promising 1, inaction 0.
4. method according to claim 1, it is characterised in that based on mounted one or more applications of user and described Similarity list, determine the user in preset application library each application preference the step of in, the user is to certain The preference of application is the mounted multiple cumulative sums applied respectively with the similarity between described certain application of user, meter Calculate formula as follows：
P l = Σ k = 1 T S k , l
Wherein：PlRepresent preference of the user to the application l in preset application library；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and preset application library.
5. method according to claim 2, it is characterised in that described pre- to recommend application be have 1 mounted with user Or the application of the label same label of multiple applications.
6. a kind of that the device applied to recommend to apply has been installed based on user, it includes：
The similarity list determining unit of application, for determining the similarity list between applying two-by-two in preset application library；
User to apply preference determining unit, for based on user it is mounted one or more application and the similarity List, determines preference of the user to each application in preset application library；
Recommendation unit, for choosing a number of corresponding from preset application library based on the preference angle value descending order Applied using as recommendation.
7. device according to claim 6, it is characterised in that the similarity list determining unit of the application is used to calculate Similarity value between applying two-by-two in preset application library, the Similarity value between these are applied two-by-two is made list, institute The computational methods for stating similarity between applying two-by-two are as follows：
S k , l = Σ j = 1 m ( r j * t k , j * t l , j ) Σ j = 1 m ( r j * t k , j ) k ≠ l
Wherein：K=1,2 ..., n j=1,2 ..., n
Sk,lRepresent the similarity between application k and application l；
N represents the number of applications in preset application library；
M represents the number of labels in preset tag set；
tk,jRepresent using whether k has label j, promising 1, inaction 0；
tl,jRepresent using whether l has label j, promising 1, inaction 0；
rjThe resolution ratio of label j is represented, wherein：
r j = 1 Σ i = 1 n t i , j
N represents the number of applications in preset application library；
ti,jRepresent using whether i has label j, promising 1, inaction 0.
8. device according to claim 6, it is characterised in that the user is used for the preference determining unit applied will The mounted multiple applications of user respectively with the cumulative sum of the similarity between the application as the user to the application Preference, computing formula is as follows：
P l = Σ k = 1 T S k , l
Wherein：PlRepresent preference of the user to the application l in preset application library；
T represents that user has installed the quantity of application；
Sk,lRepresent that user has been installed using the similarity between the application l in k and preset application library.
9. a kind of that the device applied to recommend to apply has been installed based on user, it includes：
The similarity list determining unit of application, for determining the similarity list between applying two-by-two in preset application library；
The pre- Candidate Set for recommending application sets up unit, recommends to answer as pre- for selecting multiple application from preset application library With foundation includes the plurality of pre- Candidate Set for recommending to apply；
User to apply preference determining unit, for based on user it is mounted one or more application and the similarity List, determines the user preference for recommending to apply pre- to each in the Candidate Set；
Recommendation unit, for choosing a number of corresponding from the Candidate Set based on the preference angle value descending order Applied using as recommendation.
10. device according to claim 9, it is characterised in that described pre- to recommend application be have mounted with user 1 The application of the label same label of individual or multiple applications.
CN201710072980.XA 2017-02-10 2017-02-10 Method and device for recommending applications based on applications installed by user Active CN106844724B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710072980.XA CN106844724B (en) 2017-02-10 2017-02-10 Method and device for recommending applications based on applications installed by user

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710072980.XA CN106844724B (en) 2017-02-10 2017-02-10 Method and device for recommending applications based on applications installed by user

Publications (2)

Publication Number Publication Date
CN106844724A true CN106844724A (en) 2017-06-13
CN106844724B CN106844724B (en) 2020-10-16

Family

ID=59121804

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710072980.XA Active CN106844724B (en) 2017-02-10 2017-02-10 Method and device for recommending applications based on applications installed by user

Country Status (1)

Country Link
CN (1) CN106844724B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108446311A (en) * 2018-02-06 2018-08-24 微梦创科网络科技（中国）有限公司 A kind of APP recommendation method and devices based on social networks
CN108734556A (en) * 2018-05-18 2018-11-02 广州优视网络科技有限公司 Recommend the method and device of application
CN109040164A (en) * 2018-05-21 2018-12-18 广州优视网络科技有限公司 Using recommended method, device, storage medium and computer equipment
WO2020015112A1 (en) * 2018-07-20 2020-01-23 平安科技（深圳）有限公司 Product function recommendation method, terminal device and computer-readable storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455522A (en) * 2012-06-04 2013-12-18 北京搜狗科技发展有限公司 Recommendation method and system of application extension tools
CN103677866A (en) * 2012-09-05 2014-03-26 北京搜狗科技发展有限公司 Application program extension tool pushing method and system
US20150154310A1 (en) * 2011-11-30 2015-06-04 At&T Intellectual Property I, L.P. Methods, Systems, And Computer Program Products For Recommending Applications Based On User Interaction Patterns
CN104808983A (en) * 2015-03-19 2015-07-29 深圳市梦域科技有限公司 Application program push method and server
CN106202328A (en) * 2016-07-01 2016-12-07 中国传媒大学 A kind of recommendation method for new projects' cold start-up

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150154310A1 (en) * 2011-11-30 2015-06-04 At&T Intellectual Property I, L.P. Methods, Systems, And Computer Program Products For Recommending Applications Based On User Interaction Patterns
CN103455522A (en) * 2012-06-04 2013-12-18 北京搜狗科技发展有限公司 Recommendation method and system of application extension tools
CN103677866A (en) * 2012-09-05 2014-03-26 北京搜狗科技发展有限公司 Application program extension tool pushing method and system
CN104808983A (en) * 2015-03-19 2015-07-29 深圳市梦域科技有限公司 Application program push method and server
CN106202328A (en) * 2016-07-01 2016-12-07 中国传媒大学 A kind of recommendation method for new projects' cold start-up

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108446311A (en) * 2018-02-06 2018-08-24 微梦创科网络科技（中国）有限公司 A kind of APP recommendation method and devices based on social networks
CN108734556A (en) * 2018-05-18 2018-11-02 广州优视网络科技有限公司 Recommend the method and device of application
CN109040164A (en) * 2018-05-21 2018-12-18 广州优视网络科技有限公司 Using recommended method, device, storage medium and computer equipment
CN109040164B (en) * 2018-05-21 2021-11-26 阿里巴巴（中国）有限公司 Application recommendation method and device, storage medium and computer equipment
WO2020015112A1 (en) * 2018-07-20 2020-01-23 平安科技（深圳）有限公司 Product function recommendation method, terminal device and computer-readable storage medium

Also Published As

Publication number Publication date
CN106844724B (en) 2020-10-16

Similar Documents

Publication Publication Date Title
US11698932B2 (en) Media content item recommendation system
CN106846094A (en) A kind of method and apparatus for recommending application message based on application has been installed
CN103823908B (en) Content recommendation method and server based on user preference
CN107330750B (en) A kind of recommended products figure method and device, electronic equipment
CN106844724A (en) It is a kind of that the method and apparatus applied to recommend application have been installed based on user
CN111461841B (en) Article recommendation method, device, server and storage medium
CN105095219B (en) Micro-blog recommendation method and terminal
CN105975472A (en) Method and device for recommendation
CN110245301A (en) A kind of recommended method, device and storage medium
CN106204063A (en) A kind of paying customer's method for digging and device
CN105335409A (en) Target user determination method and device and network server
CN106168980A (en) Multimedia resource recommends sort method and device
CN104751354B (en) A kind of advertisement crowd screening technique
CN107729578B (en) Music recommendation method and device
CN104133817A (en) Online community interaction method and device and online community platform
CN109409928A (en) A kind of material recommended method, device, storage medium, terminal
CN106909688A (en) A kind of method and apparatus that search word is recommended based on input search word
CN112860937B (en) KNN and word embedding based mixed music recommendation method, system and equipment
CN106874503B (en) Method and device for acquiring recommended data
CN106951571A (en) A kind of method and apparatus for giving application mark label
CN106980667B (en) A kind of method and apparatus to article mark label
CN110351318A (en) Using the method, terminal and computer storage medium of recommendation
CN107683473A (en) The automatic playlist generation of properties collection
CN104820689B (en) It is a kind of that single method and system associated with personal account is sung into Karaoke
CN107562848A (en) A kind of video recommendation method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20200902

Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province

Applicant after: Alibaba (China) Co.,Ltd.

Address before: 510627 Guangdong city of Guangzhou province Whampoa Tianhe District Road No. 163 Xiping Yun Lu Yun Ping square B radio tower 15 layer self unit 02

Applicant before: GUANGZHOU UC NETWORK TECHNOLOGY Co.,Ltd.

GR01 Patent grant
GR01 Patent grant