CN108304428A - Information recommendation method and device - Google Patents
Information recommendation method and device Download PDFInfo
- Publication number
- CN108304428A CN108304428A CN201710299871.1A CN201710299871A CN108304428A CN 108304428 A CN108304428 A CN 108304428A CN 201710299871 A CN201710299871 A CN 201710299871A CN 108304428 A CN108304428 A CN 108304428A
- Authority
- CN
- China
- Prior art keywords
- group
- information
- recommendation
- account number
- scoring
- 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.)
- Pending
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
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/74—Browsing; Visualisation therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/735—Filtering based on additional data, e.g. user or group profiles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Human Computer Interaction (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of information recommendation method and devices, belong to field of computer technology.The method includes:Determine the group that target account number is added;According to the historical information that the group member in each group was checked, the information recommendation list corresponding to each group is determined;According to described information recommendation list to the target account number recommendation information.It solves in the prior art when target account number is the account number of new registration, server can not send recommendation information according to the historical viewings information of the target account number to the target account number, namely server can only send the recommendation information of acquiescence to target account number at this time, the relatively low problem of the accuracy rate of the recommendation information of transmission;Even if it is new account to have reached target account number, server still can send recommendation information for the target account number according to the group that account is added, improve the effect of the accuracy rate of the information of recommendation.
Description
Technical field
The present embodiments relate to field of computer technology, more particularly to a kind of information recommendation method and device.
Background technology
In video playing client, server can be to the possible interested video of user recommended user.
Existing video recommendation method includes:Server obtains the history viewing record of target account number;According to the conception of history
See that record recommends relevant video to the target account number.For example, recommending same type of video in being recorded with history viewing, such as push away
Recommend funny video.
If target account number is the account number of new registration, the content in history viewing record may be seldom namely this is gone through
History viewing record can not characterize the interest of user, therefore server is recorded as target account number recommendation according to history viewing and regards
When frequency, the interested video of video not necessarily user of recommendation, namely the accuracy rate of the video of recommendation are relatively low.
Invention content
In order to solve the problems in the existing technology, an embodiment of the present invention provides a kind of information recommendation method and dresses
It sets.Technical solution is as follows:
According to a first aspect of the embodiments of the present invention, a kind of information recommendation method is provided, this method includes:
Determine the group that target account number is added;
According to the historical information that the group member in each group was checked, the information recommendation corresponding to each group is determined
List;
According to described information recommendation list to the target account number recommendation information.
According to a second aspect of the embodiments of the present invention, a kind of information recommending apparatus is provided, this method includes:
Group determination module, the group being added for determining target account number;
List determining module, the historical information for being checked according to the group member history in each group determine every
Information recommendation list corresponding to a group;
Information sending module, described information recommendation list for being determined according to the list determining module is to the target
Account number recommendation information.
The advantageous effect that technical solution provided in an embodiment of the present invention is brought is:
The group being added by determining target account number, and then the historical information next life checked according to the group member in group
At the information recommendation list of the group, according to each information recommendation list to target account number recommendation information;Solves the prior art
In when target account number is the account number of new registration, server can not be according to the historical viewings information of the target account number to the target account
Number recommendation information is sent, namely server can only send the recommendation information of acquiescence to target account number, the recommendation information of transmission at this time
The relatively low problem of accuracy rate;Even if it is new account to have reached target account number, server still can be added according to account
Group come for the target account number send recommendation information, improve the effect of the accuracy rate of the information of recommendation.
Description of the drawings
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for
For those of ordinary skill in the art, without creative efforts, other are can also be obtained according to these attached drawings
Attached drawing.
Fig. 1 is the schematic diagram of the implementation environment involved by each embodiment of the present invention;
Fig. 2 is the flow chart of information recommendation method provided by one embodiment of the present invention;
Fig. 3 is the flow chart provided by one embodiment of the present invention for sending recommendation information;
Fig. 4 is the block diagram of information recommendation method provided by one embodiment of the present invention;
Fig. 5 is the schematic diagram of the performance for the information that different information recommendation methods provided by one embodiment of the present invention are recommended;
Another signal of the performance for the information that Fig. 6 different information recommendation methods provided by one embodiment of the present invention are recommended
Figure;
Fig. 7 is the schematic diagram of information recommending apparatus provided by one embodiment of the present invention;
Fig. 8 is the schematic diagram of server provided by one embodiment of the present invention.
Specific implementation mode
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Formula is described in further detail.
Referring to FIG. 1, it illustrates the schematic diagrames of the implementation environment involved by each embodiment of the present invention, as shown in Figure 1,
The implementation environment may include terminal 110 and server 120.
Terminal 110 can log in target account number.In actual implementation, client can have been run in terminal 110, for example, can
There are video playing client, audio client end of playing back or news client with operation, and mesh is logged in the client of operation
Mark account number;Optionally, terminal 110 can also log in the target account number in webpage, and the present embodiment does not limit this.The mesh
Mark the account number that account number is the social networking application client for supporting group management function, and the social networking application client and above-mentioned described visitor
Family end can be same client, or different clients do not limit this.In addition, the group described in the present embodiment
Group can be the discussion group that the group on ordinary meaning may be more people's chats, not limit this.Terminal 110 can be
Such as terminal of mobile phone, tablet computer, electronic reader, desktop computer or laptop etc does not limit this.
Terminal 110 can be connect by wired or wireless network with server 120.
Server 120 is used to the background server that terminal 110 provides background service, which can be one
Platform server, or the server cluster being made of multiple servers.For example, logging in the mesh in webpage with terminal 110
For marking account number, for such situation, server 120 may include the backstage clothes corresponding to the search client in terminal 110
The server corresponding to social networking application client corresponding to business device and target account number.In actual implementation, server 120 has
The ability of recommendation information is sent from trend terminal 110 by the target account number that is logged in terminal 110.
Referring to FIG. 2, it illustrates the method flow diagram of information recommendation method provided by one embodiment of the present invention, this reality
Example is applied with the information recommendation method for being illustrated in server 120 shown in FIG. 1.As shown in Fig. 2, the information recommendation side
Method may include:
Step 201, the group that target account number is added is determined.
Can be stored with the description information of each account number in server, the description information include account number friend relation chain,
Group that account number is added, account number ID (Identification, identity), the account number pet name etc., therefore, server can be from
The group of target account number addition is got in the description information of target account number.
Optionally, server can receive target account number reach the standard grade request when, determine target account number be added group;
Alternatively, the group that server can be added with the target account number that timing determination is online, for example, daily 16:30 determine
The group that each target account number of line is added;Or it is asked in the opening for receiving the default interface of the opening from target account number
When asking, the group that target account number is added is determined, for example, when receiving the opening request of target account number opening information interface, determine
The group that target account number is added.Certainly in actual implementation, server can also execute under other scenes and determine that target account number adds
The step of group entered, the present embodiment does not limit this.
For the present embodiment by taking server recommends video to video playing client as an example, server can be in video playing client
When end is logged in by target account number, the group that the target account number is added is determined.
Step 202, the recommendation index for the historical information that the group member in each group was checked is obtained.
Historical information refers to the historical information that the group member history in group was checked, which can be group
The information that the server that member checked is recommended, or the information oneself actively searched for checked does not limit this
It is fixed.And in actual implementation, which can be video information, audio-frequency information or other types of information.Following realities
It applies example unless otherwise specified, using information as video information, namely recommends video to illustrate to target account number.
Optionally, this step may include:
First, the recommendation factor of each historical information in each group is obtained, it includes the group of historical information to recommend the factor
The discrimination of interior temperature and/or historical information.Wherein, temperature is used to indicate the journey that the historical information is checked in group in group
Degree, discrimination is for indicating that the historical information distinguishes the degree of other historical informations.
To recommend the factor to include in the group of historical information for temperature, the step of server obtains temperature in group, includes:It is right
The number that the group member of the historical information is checked in each group, server statistics group, the number that statistics is obtained are made
For temperature in the group of the historical information.For example, temperature in groupWherein, ηk,jIndicate video j in group
Temperature in group in k, i indicate a group member in group k, UkIndicate the set of the group member in group k, ViIt indicates
The video of target account number checks record.I(j∈Vi) checked by group member i for 1 expression video j, and I (j ∈ Vi) it is 0 expression
Video j was not checked by group member i.
For recommending the discrimination that the factor includes historical information, server obtain discrimination the step of include:According to having
Group member checked that the number of the group of historical information and the total number of group calculated discrimination.Wherein, the sum of group
Refer to the sum of all groups in the system corresponding to target account number, discrimination and the negatively correlated pass of target ratio of historical information
System.Target ratio be have group member checked historical information group number and group sum ratio.For example, distinguishing
DegreeWherein, G is the sum of group in system, I (j ∈ Pk) be 1 indicate video j by k groups
At least one user checked in group, I (j ∈ Pk) it is that 0 expression video j was not checked that g was indicated by any user in k groups
The set of all groups, P in systemkIndicate the video pond of group k.
And when recommending the factor simultaneously including above-mentioned the two, server can obtain history respectively by above-mentioned acquisition methods
Temperature and discrimination in the group of information, details are not described herein.Also, it includes both above-mentioned that the present embodiment, which is also to recommend the factor,
At least one of illustrate, optionally, it can also includes other content to recommend the factor, do not limited this.
The present embodiment for being obtained by above-mentioned acquisition methods and recommend the factor, in actual implementation, can also only pass through
Other methods obtain, and the present embodiment does not limit this.
Since each group member in group may not check any information, this step can in actual implementation
It can be not present, details are not described herein.
Second, the recommendation index of historical information is calculated according to the recommendation factor of every historical information.
Optionally, if the factor is recommended to only include one kind, the recommendation factor the pushing away as the historical information that will be calculated
Recommend index.
And if recommend the factor include at least two, server according at least two recommendation the factors calculate the recommendation index.
For example, recommending index Wk,j=ηk,j*βj.Optionally, server can be according to each value for recommending the factor and each recommendation
The weight of the factor calculates the recommendation index of the historical information.Wherein, it can be pre- in server each to recommend the weight of the factor
The weight first set, or user is self-defined in advance and stores the weight into server, does not limit this.
In actual implementation, before the recommendation index for obtaining historical information, server can be according to default filtering rule mistake
Filter the historical information that the group member history in each group is checked.Optionally, server, which can filter, checks that duration is more than the
The information of one preset duration.First preset duration can be preset duration in server, can also be server root
The duration that duration calculation used when the information obtains is checked according to each user, for example, each user is checked the letter by server
The average value of duration used does not limit this as first preset duration when breath.Certainly it is video information in the information
Either first preset duration can also be video information or the duration of audio-frequency information when audio-frequency information, and the present embodiment is to this
It does not limit.Optionally, server can also filter the information for checking that duration is less than the second preset duration.Wherein, second is pre-
If duration can be preset duration in server, or user-defined duration does not limit this.By
When checking that duration is more than the first preset duration, illustrate that the information is read extremely in information, therefore by being carried out to the information
Filtering ensure that the accuracy subsequently calculated.Similarly, since information is when checking that duration is less than the second preset duration, illustrate the letter
Breath may be clicked by falseness, can not think that group member had checked the information at this time, therefore by being carried out to the information
Filtering, ensure that the accuracy subsequently calculated.Optionally, server, which can also filter, checks that the total quantity of information is more than default threshold
The historical information that the group member of value is checked.When due to checking that the total quantity of information is more than predetermined threshold value, illustrate that account can
Can be not manual operation, it is likely to web crawlers, therefore the information by being checked to such group member is filtered, and is protected
The accuracy of follow-up calculating is demonstrate,proved.
Step 203, the letter corresponding to each group is generated according to the recommendation index of each historical information in each group
Cease recommendation list.
Optionally, after the recommendation index of each historical information during each group is calculated, server can root
Ranking is carried out according to each historical information of each recommendation exponent pair, the list that ranking is obtained is arranged as the information recommendation of the group
Table.Certainly, in actual implementation, server can select wherein ranking in preceding n of information, and each history that selection is obtained
Information recommendation list of the list of information as the group.Optionally, server is also an option that it is more than default threshold to recommend index
Each historical information of value, and using each historical information selected as the information recommendation list of the group.
Step 204, according to information recommendation list to target account number recommendation information.
After determining and obtaining the information recommendation list corresponding to each group, server is to target account number recommendation information.
Optionally, server can send the information in each information recommendation list to target account number, alternatively, server can be sent respectively
Each information in a information recommendation list links to target account number, and the present embodiment does not limit this.
In actual implementation, the item number of the recommendation information of transmission can be default number of branches, at this point, server can be from each letter
The information for selecting default number of branches in recommendation list is ceased, and then recommends the information that selection obtains to target account number.For example, to preset item
Number is 5, and for the number of information recommendation list is 5, server can select one to believe from each information recommendation list
It ceases, and recommends the information of selection to target account number.Optionally, if the ratio of default number of branches and the number of information recommendation list is not
Integer, then server can be to each information recommendation list ordering, and selects a letter from each information recommendation list successively
Cease and then finally obtain the information of default number of branches.For example, still by taking default number of branches are 5 as an example, it is assumed that information recommendation list
Number is 3, then at this point, server can be successively from first information recommendation list, second information recommendation list, third
An information is selected respectively in information recommendation list, first information recommendation list and second information recommendation list, to mesh
Mark account number recommends 5 information that selection obtains.
Optionally, before step 201, server can also model target account number, detect the target account number
Whether registration time length is more than preset duration, if the target account number is the account that registration time length is less than preset duration, server is true
The fixed target account number is new account, at this point, the step of server can execute the group that determining target account number is added;And if target
Account number is that registration time length is no less than the account number of preset duration, then server can detect the history of the target account number and check note at this time
Whether the item number of record is more than default number of branches, if being more than, illustrate the history of the target account number check record it is enough, at this time can be with
Checked according to history and be recorded as the target account number recommendation information, and if history check record item number be no more than preset condition,
Illustrate to check that the accuracy rate for being recorded as recommending when target account number is recommended may be relatively low according to history at this time, at this point, server can be with
Execute the step of determining the group that target account number is added.
In an application scenarios of the present embodiment, after video playing client is logged in using target account number, video
The background server of client end of playing back can determine the target account number be added group (background server can be with linking objective account number
Social networking application client corresponding to social networking application server, and from the social networking application server obtain target account number be added
Group), determining target account number be added group include:Family crowd, colleague group and university classmate group;Obtain the group in group
The recommendation index for the historical information that group membership's history was checked;It is given birth to according to the recommendation index of each historical information in each group
At the information recommendation list of the group, for example, for family crowd, information recommendation list that background server determines include video A,
Video B and video C, for the group that works together, the information recommendation list that background server determines includes video A, video C and video E;It is right
In university classmate group, the information recommendation list that background server determines includes video A, video B, video D, video E and video F;
Background server recommends recommendation information according to determining information recommendation list to target account number, for example, background server can be to
Target account number recommends video A, video B, video C, video D, video E and video F.
In conclusion information recommendation method provided in this embodiment, the group being added by determining target account number, Jin Ergen
The historical information checked according to the group member history in group generates the information recommendation list of the group, is pushed away according to each information
List is recommended to target account number recommendation information;Solve in the prior art target account number be new registration account number when, server without
Method sends recommendation information according to the historical viewings information of the target account number to the target account number, namely server can only be sent at this time
The recommendation information of acquiescence is to target account number, the relatively low problem of the accuracy rate of the recommendation information of transmission;Even if having reached target account number
For new account, server still can send recommendation information for the target account number according to the group that account is added, improve
The effect of the accuracy rate for the information recommended.
In the above-described embodiments, when the group that target account number is added has at least two, referring to FIG. 3, above-mentioned steps 204
May include:
Step 204a obtains the scoring of each group.
This step may include:
First, obtain group's scoring factor of each group, group's scoring factor includes group's liveness, social similar
At least one of degree, interest liveness and Interest Similarity.Wherein:
Group's liveness is used to indicate the active degree of the group member in group.In actual implementation, server can root
It is calculated according to the total quantity of the group message of group member transmission in historical time section and the quantity of the group member in group
Group's liveness.For example, group's livenessWherein, Mk,iIndicate that user i is in group in historical time section
The total quantity of the information sent in group k, | uk| indicate the quantity of the group member in group k.
Social similarity is used to indicate the similarity degree of the group member social activity in group.In actual implementation, server can
With according to the total quantity of the group member in the quantity of the group member of good friend and group calculates the social activity phase each other in group
Like degree.For example, social similarityWherein, Iu,iBe according to social account number u and social account number i whether be
The quantity of good friend two-by-two in the group that the mark of good friend is calculated, | Uk| indicate the quantity of group member in group k.
Interest liveness is used to indicate the probability that information is checked in group.In actual implementation, server can be according to letter
The recommendation index of breath calculates the interest liveness.For example, interest livenessWherein, Wk,jTo recommend index,
And Wk,jCalculation it is similar with the calculation in the step 202 in above-described embodiment, details are not described herein.
Interest Similarity is used to indicate the similarity of group member historical behavior.In actual implementation, server can be according to group
The Interest Similarity of interior any two group member calculates the Interest Similarity of the group.For example, any two group forms in group
The Interest Similarity of memberThe then Interest Similarity of the group
Wherein, u and i is two group members in group k, VuIt is recorded for the historical operation of group member u, ViFor going through for group member i
History operation note, UkFor the quantity of the group member in group k.Historical operation record may include check information check note
Record, the subscription behavior etc. of the review record of comment information, subscription information, details are not described herein.
It is above-mentioned only to include at least one of above-mentioned four kinds with group's scoring factor and come for example, optionally, group
The factor that scores can also include other content, and the present embodiment does not limit this.
Second, the scoring for factor calculating group of being scored according to the group of each group.
Optionally, server can be according to the weight corresponding to each group scoring factor and each group scoring factor
The scoring for calculating the group does not limit this.Certainly in actual implementation, server can also according to group score the factor with
And logistic regression calculates the scoring of the group, details are not described herein.
For the present embodiment is the scoring by server by the acquisition group scoring factor to calculate group, optionally,
Server can also obtain pre-stored scoring, and the pre-stored scoring of server is server previously according to the group got
The scoring that the group scoring factor is calculated and preserved, details are not described herein.
Step 204b determines the letter for needing to recommend according to the scoring of each group and the information recommendation list of each group
Breath.
Optionally, this step may include:
First, for every information in each information recommendation list, according to each information recommendation list comprising information
The scoring of corresponding group calculates the scoring of information.
Optionally, the ranking that server can be according to the score and information of corresponding group in information recommendation list
Calculate the scoring of the information.For example, for information A, first in information recommendation list 1, information recommendation list 3
Include information A in 4, the 2nd of information recommendation list 4, then scoring * 0.4+ groups 3 of scoring=group 1 of the information
Scoring * 0.2+ groups 4 scoring * 0.1.
Optionally, server can also refer to according to recommendation of the score and the information of corresponding group in the group
It counts to calculate the scoring of the information.Its computational methods with it is above-mentioned similar according to the computational methods of ranking, details are not described herein.
Certainly, in actual implementation, the score of each information, this implementation can also be calculated in server by other means
Example does not limit this.
After above-mentioned steps, the scoring of each information can be calculated in server, such as 10 information score by
It is high to Low be respectively 98,90,82 ..., 76.
Second, the information for needing to recommend is determined according to the scoring for each information being calculated.
Optionally, server can select the information that the information for being scored above preset fraction is recommended as needs;Alternatively, clothes
Being engaged in device can be according to the sequence progress ranking of scoring from high to low, and wherein ranking is selected to be pushed away as needs in preceding n of information
The information recommended.Certainly, in actual implementation, server can also determine the information that needs are recommended, this implementation by other means
Example does not limit this.
In actual implementation, after the scoring of group is calculated, server is also an option that the group for meeting preset condition
Group, and then the scoring of group is obtained according to selection and the information recommendation list of the group determines information that needs are recommended, this
Details are not described herein for embodiment.Wherein, the group for meeting preset condition can be group of the scoring higher than preset fraction or comment
Divide ranking in preceding m of group, details are not described herein.
Step 204c recommends determining information to target account number.
After determining the information that needs are recommended, server can recommend determining letter to the terminal corresponding to target account number
Breath.Correspondingly, terminal can receive the information by the target account number.
Referring to FIG. 4, the complete block diagram of the information recommendation method provided it illustrates above-described embodiment, as shown in figure 4,
It, can be according to hot in the group for the historical information that the group member history in each group was checked after determining the group being added
Degree and discrimination calculate the recommendation index of the historical information, and then obtain the information recommendation list of each group.At the same time it can also
According to group score the factor namely interest liveness shown in figure, Interest Similarity, social liveness and social similarity come
Calculate the scoring of each group;Qualified group is filtered out according to the scoring of each group, and then obtained according to screening
The scoring of group and the information recommendation list of the group merge to obtain the information for needing to recommend.In the information for determining that needs are recommended
Later, can also be that determining information adds description information.As shown in figure 4, before determining the group that target account number is added, clothes
Business device can also first model target account number, if the target account number is the account that registration time length is less than preset duration, take
Business device determines that the target account number is new account, at this point, the step of server can execute the group that determining target account number is added;And
If target account number is the account number that registration time length is no less than preset duration, server can detect the history of the target account number at this time
Check record item number whether be more than default number of branches, if being more than, illustrate the history of the target account number check record it is enough, this
When can be checked according to history and be recorded as the target account number recommendation information, and if history checks that the item number of record is no more than default item
Part then illustrates to check that the accuracy rate for being recorded as recommending when target account number is recommended may be relatively low according to history at this time, at this point, server
The step of group that determining target account number is added can be executed.
In order to weigh the accuracy of information recommendation method provided in this embodiment, it can be tested by A/B and it is surveyed
Examination, in A/B tests, test user is evenly distributed to several set at random, and each set is special using a specific target
The setting of property, keeps other settings identical (similar control quantity method).It is mutual that these collection share one group of predefined evaluation index
Compare.In the present embodiment, target property is using different algorithms.Control group algorithm:Collaborative filtering, content-based recommendation
And the information recommendation method described in above-described embodiment.
In order to weigh the performance for comparing these algorithms in correlation and diversity etc., the present embodiment has used two fingers
Mark, i.e. CTR (Click Through Rate, clicking rate) and Gini coefficients.
Wherein, " #of impression " indicates to recommend the quantity of the information of user, and " #of click " expression is recommended
The quantity for the information being clicked after user information.CTR is bigger, and it is more accurate to indicate to recommend.
Wherein, n represents the quantity of recommended information, djIt is to be recommended number ranking after ranking from high to low according to information
Recommendation number in jth position.Gini coefficients are smaller, indicate that recommendation results have higher diversity.
Because of data sensitive, the second largest CTR values are normalized to 1 by the present embodiment, the CTR values of other two algorithms according to
Corresponding variation is done with its ratio.The result for normalizing CTR values and Gini coefficients is as shown in Figure 5 and Figure 6, wherein each value is
Daily average result.
In the above-described embodiments, when to target account number recommendation information, server can also be every information setting description
Information.The description information is for indicating that the information was checked by the group member of which group.For example, for information A, it should
Information A is checked by the group member in group A and group B, then at this point, server is when recommending information A to target account number, clothes
Business device can also send description information that content is " information A is checked by the user in group A and group B " to target account number.
By based on group come recommendation information so that for information be arranged description information after, can both pass through setting
Description information checks the probability of the information to improve user, and can guarantee that the group member in group checks the privacy peace of information
Entirely.
The code of above- mentioned information proposed algorithm is by HIVE (Tool for Data Warehouse based on Hadoop) SQL
(Structured Query Language, structured query language) and python (the explanation type computer journeys of object-oriented
Sequence design language) it writes.HIVE can write code with class SQL statement, and SQL statement is converted to MapReduce automatically and is appointed
Business is run.HIVE SQL are responsible for part data preparation, the generation of information recommendation list and the determining information for needing to recommend;
Python codes are mainly responsible for this spy of logic in the scoring for calculating group and return part and routine execution part.
Referring to FIG. 7, it illustrates the structural schematic diagram of information recommending apparatus provided by one embodiment of the present invention, such as scheme
Shown in 7, which may include:Group determination module 710, list determining module 720 and information sending module
730。
Group determination module 710, the group being added for determining target account number;
List determining module 720, the historical information for being checked according to the group member in each group determine each
Information recommendation list corresponding to group;
Information sending module 730, described information recommendation list for being determined according to the list determining module 720 is to institute
State target account number recommendation information.
In conclusion information recommending apparatus provided in this embodiment, the group being added by determining target account number, Jin Ergen
The historical information checked according to the group member history in group generates the information recommendation list of the group, is pushed away according to each information
List is recommended to target account number recommendation information;Solve in the prior art target account number be new registration account number when, server without
Method sends recommendation information according to the historical viewings information of the target account number to the target account number, namely server can only be sent at this time
The recommendation information of acquiescence is to target account number, the relatively low problem of the accuracy rate of the recommendation information of transmission;Even if having reached target account number
For new account, server still can send recommendation information for the target account number according to the group that account is added, improve
The effect of the accuracy rate for the information recommended.
Based on the information recommending apparatus that above-described embodiment provides, optionally, the list determining module 720, including:
Index acquiring unit, the recommendation index for obtaining the historical information that the group member in each group was checked;
List generation unit, for generating each institute of group according to the recommendation index of each historical information in each group
Corresponding information recommendation list.
Optionally, the index acquiring unit, including:
First obtain subelement, the recommendation factor for obtaining each historical information in each group, the recommendation because
Attached bag includes the discrimination of temperature and/or the historical information in the group of the historical information;
Index computation subunit is pushed away for being obtained according to described first described in every historical information that subelement is got
Recommend the recommendation index that the factor calculates the historical information.
Optionally, it is described first obtain subelement, be additionally operable to the group being added according to the target account number number and
There is group member to check the number of the group of the historical information, calculates the discrimination.
Optionally, described information sending module 730, including:
Score acquiring unit, for when the group that the target account number is added includes at least two, obtaining each group
Scoring;
Information determination unit, the scoring of each group for being got according to the scoring acquiring unit and each group
The information recommendation list of group determines the information for needing to recommend;
Information transmitting unit, for recommending determining described information to the target account number.
Optionally, the scoring acquiring unit, including:
Second obtains subelement, and group's scoring factor for obtaining each group, group's scoring factor includes group
At least one of group liveness, social similarity, interest liveness and Interest Similarity;
Score computation subunit, for obtained according to described second group's scoring of each group for getting of subelement because
Son calculates the scoring of the group.
Optionally, described information determination unit is additionally operable to:
For every information in each information recommendation list, according to each information recommendation list institute comprising described information
The scoring of corresponding group calculates the scoring of described information;
The information for needing to recommend is determined according to the scoring for each information being calculated.
It should be noted that:The information recommending apparatus that above-described embodiment provides, only being partitioned into above-mentioned each function module
Row can be completed by different function modules as needed and by above-mentioned function distribution for example, in practical application, i.e., will clothes
The internal structure of business device is divided into different function modules, to complete all or part of the functions described above.In addition, above-mentioned
The information recommending apparatus and information recommendation method embodiment that embodiment provides belong to same design, the specific implementation process side of referring to
Method embodiment, which is not described herein again.
The embodiment of the present invention additionally provides a kind of computer readable storage medium, which can be
Computer readable storage medium included in memory;Can also be individualism, without the computer in supplying server
Readable storage medium storing program for executing.The computer-readable recording medium storage there are one either more than one program this or one with
Upper program is used for executing above- mentioned information by one or more than one processor and recommends method.
Referring to FIG. 8, it illustrates the structural schematic diagrams of server provided by one embodiment of the present invention.The server is used
In the information recommendation method for implementing to provide in above-described embodiment.Specifically:
The server 800 is including central processing unit (CPU) 801 including random access memory (RAM) 802 and only
Read the system storage 804 of memory (ROM) 803, and the system of connection system storage 804 and central processing unit 801
Bus 805.The server 800 further includes the basic input/output of transmission information between each device helped in computer
System (I/O systems) 806, and large capacity for storage program area 813, application program 814 and other program modules 815 are deposited
Store up equipment 807.
The basic input/output 806 includes display 808 for showing information and inputs letter for user
The input equipment 809 of such as mouse, keyboard etc of breath.The wherein described display 808 and input equipment 809 are all by being connected to
The input and output controller 810 of system bus 805 is connected to central processing unit 801.The basic input/output 806
Can also include input and output controller 810 for receive and handle from keyboard, mouse or electronic touch pen etc. it is multiple its
The input of his equipment.Similarly, input and output controller 810 also provides output to display screen, printer or other kinds of defeated
Go out equipment.
The mass-memory unit 808 is by being connected to the bulk memory controller (not shown) of system bus 805
It is connected to central processing unit 801.The mass-memory unit 807 and its associated computer-readable medium are server
800 provide non-volatile memories.That is, the mass-memory unit 808 may include such as hard disk or CD-ROM
The computer-readable medium (not shown) of driver etc.
Without loss of generality, the computer-readable medium may include computer storage media and communication media.Computer
Storage medium includes information such as computer-readable instruction, data structure, program module or other data for storage
The volatile and non-volatile of any method or technique realization, removable and irremovable medium.Computer storage media includes
RAM, ROM, EPROM, EEPROM, flash memory or other solid-state storages its technologies, CD-ROM, DVD or other optical storages, tape
Box, tape, disk storage or other magnetic storage apparatus.Certainly, skilled person will appreciate that the computer storage media
It is not limited to above-mentioned several.Above-mentioned system storage 804 and mass-memory unit 808 may be collectively referred to as memory.
According to various embodiments of the present invention, the server 800 can also be arrived by network connections such as internets
Remote computer operation on network.Namely server 800 can be by the network interface that is connected on the system bus 805
Unit 811 is connected to network 812, in other words, can also be connected to using Network Interface Unit 811 other kinds of network or
Remote computer system (not shown).
The memory further includes that one or more than one program, the one or more programs are stored in
In memory, and it is configured to be executed by one or more than one processor.Said one or more than one program include
Recommend the instruction of method for executing above- mentioned information.
It should be understood that it is used in the present context, unless context clearly supports exception, singulative " one
It is a " (" a ", " an ", " the ") be intended to also include plural form.It is to be further understood that "and/or" used herein is
Finger includes one or the arbitrary and all possible combinations of more than one project listed in association.
The embodiments of the present invention are for illustration only, can not represent the quality of embodiment.
One of ordinary skill in the art will appreciate that realizing that all or part of step of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can be stored in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent replacement, improvement and so on should all be included in the protection scope of the present invention.
Claims (14)
1. a kind of information recommendation method, which is characterized in that the method includes:
Determine the group that target account number is added;
According to the historical information that the group member in each group was checked, the information recommendation row corresponding to each group are determined
Table;
According to described information recommendation list to the target account number recommendation information.
2. according to the method described in claim 1, it is characterized in that, what the group member in each group of the basis was checked
Historical information determines the information recommendation list corresponding to each group, including:
Obtain the recommendation index for the historical information that the group member in each group was checked;
The information recommendation list corresponding to each group is generated according to the recommendation index of each historical information in each group.
3. according to the method described in claim 2, it is characterized in that, what the group member obtained in each group was checked
The recommendation index of historical information, including:
The recommendation factor of each historical information in each group is obtained, the recommendation factor includes in the group of the historical information
The discrimination of temperature and/or the historical information;
The recommendation index of the historical information is calculated according to the recommendation factor of every historical information.
4. according to the method described in claim 3, it is characterized in that, the area for obtaining each historical information in each group
Indexing, including:
According to there is group member to check, the number of group of the historical information and the total number of group calculate the differentiation
Degree.
5. method according to any one of claims 1 to 4, which is characterized in that it is described according to described information recommendation list to institute
Target account number recommendation information is stated, including:
If the group that the target account number is added includes at least two, the scoring of each group is obtained;
The information for needing to recommend is determined according to the scoring of each group and the information recommendation list of each group;
Recommend determining described information to the target account number.
6. according to the method described in claim 5, it is characterized in that, the scoring for obtaining each group, including:
Group's scoring factor of each group is obtained, group's scoring factor includes group's liveness, social similarity, interest
At least one of liveness and Interest Similarity;
The scoring of the group is calculated according to the group of each group scoring factor.
7. according to the method described in claim 5, it is characterized in that, scoring and each group of the basis each group
Information recommendation list determines the information for needing to recommend, including:
For every information in each information recommendation list, corresponding to each information recommendation list comprising described information
Group scoring calculate described information scoring;
The information for needing to recommend is determined according to the scoring for each information being calculated.
8. a kind of information recommending apparatus, which is characterized in that described device includes:
Group determination module, the group being added for determining target account number;
List determining module, the historical information for being checked according to the group member in each group determine each institute of group
Corresponding information recommendation list;
Information sending module, described information recommendation list for being determined according to the list determining module is to the target account number
Recommendation information.
9. device according to claim 8, which is characterized in that the list determining module, including:
Index acquiring unit, the recommendation index for the historical information that the group member history for obtaining in each group was checked;
List generation unit, for being generated corresponding to each group according to the recommendation index of each historical information in each group
Information recommendation list.
10. device according to claim 9, which is characterized in that the index acquiring unit, including:
First obtains subelement, and the recommendation factor for obtaining each historical information in each group, the recommendation is because of attached bag
Include the discrimination of temperature and/or the historical information in the group of the historical information;
Index computation subunit, for according to described first obtain the recommendation of every historical information that subelement is got because
Son calculates the recommendation index of the historical information.
11. device according to claim 10, which is characterized in that
Described first obtains subelement, be additionally operable to according to the number of the group that there is group member to check the historical information and
The total number of group calculates the discrimination.
12. according to any device of claim 8 to 11, which is characterized in that described information sending module, including:
Score acquiring unit, for when the group that the target account number is added includes at least two, obtaining commenting for each group
Point;
Information determination unit, the scoring of each group and each group for being got according to the scoring acquiring unit
Information recommendation list determines the information for needing to recommend;
Information transmitting unit, for recommending determining described information to the target account number.
13. device according to claim 12, which is characterized in that the scoring acquiring unit, including:
Second obtains subelement, and group's scoring factor for obtaining each group, group's scoring factor includes that group lives
At least one of jerk, social similarity, interest liveness and Interest Similarity;
Score computation subunit, based on the group's scoring factor for obtaining each group that subelement is got according to described second
Calculate the scoring of the group.
14. device according to claim 12, which is characterized in that described information determination unit is additionally operable to:
For every information in each information recommendation list, corresponding to each information recommendation list comprising described information
Group scoring calculate described information scoring;
The information for needing to recommend is determined according to the scoring for each information being calculated.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710299871.1A CN108304428A (en) | 2017-04-27 | 2017-04-27 | Information recommendation method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710299871.1A CN108304428A (en) | 2017-04-27 | 2017-04-27 | Information recommendation method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108304428A true CN108304428A (en) | 2018-07-20 |
Family
ID=62872178
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710299871.1A Pending CN108304428A (en) | 2017-04-27 | 2017-04-27 | Information recommendation method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108304428A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109756563A (en) * | 2018-12-17 | 2019-05-14 | 平安科技(深圳)有限公司 | Information-pushing method, device, computer equipment and storage medium |
CN110929170A (en) * | 2019-11-22 | 2020-03-27 | 腾讯科技(深圳)有限公司 | Friend recommendation method, device, equipment and storage medium for social group |
CN110941756A (en) * | 2018-09-25 | 2020-03-31 | 夏普株式会社 | Terminal device, information processing system, method, and recording medium |
CN111130992A (en) * | 2019-11-22 | 2020-05-08 | 北京达佳互联信息技术有限公司 | Group recommendation method and device, electronic equipment and storage medium |
CN111523035A (en) * | 2020-04-24 | 2020-08-11 | 上海异势信息科技有限公司 | Recommendation method, device, server and medium for APP browsing content |
CN112769676A (en) * | 2020-12-31 | 2021-05-07 | 上海掌门科技有限公司 | Method and equipment for providing information in group |
CN112989276A (en) * | 2019-12-12 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Evaluation method and device of information push system |
CN114765624A (en) * | 2020-12-31 | 2022-07-19 | 北京达佳互联信息技术有限公司 | Information recommendation method and device, server and storage medium |
CN110929170B (en) * | 2019-11-22 | 2024-06-28 | 腾讯科技(深圳)有限公司 | Friend recommendation method, device and equipment of social group and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103391302A (en) * | 2012-05-08 | 2013-11-13 | 阿里巴巴集团控股有限公司 | Information sending method and system |
CN103731738A (en) * | 2014-01-23 | 2014-04-16 | 哈尔滨理工大学 | Video recommendation method and device based on user group behavioral analysis |
CN104573109A (en) * | 2015-01-30 | 2015-04-29 | 深圳市中兴移动通信有限公司 | System, terminal and method for automatic recommendation based on group relation |
CN104572889A (en) * | 2014-12-24 | 2015-04-29 | 深圳市腾讯计算机***有限公司 | Method, device and system for recommending search terms |
CN105574045A (en) * | 2014-10-17 | 2016-05-11 | 深圳市腾讯计算机***有限公司 | Video recommendation method and server |
-
2017
- 2017-04-27 CN CN201710299871.1A patent/CN108304428A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103391302A (en) * | 2012-05-08 | 2013-11-13 | 阿里巴巴集团控股有限公司 | Information sending method and system |
CN103731738A (en) * | 2014-01-23 | 2014-04-16 | 哈尔滨理工大学 | Video recommendation method and device based on user group behavioral analysis |
CN105574045A (en) * | 2014-10-17 | 2016-05-11 | 深圳市腾讯计算机***有限公司 | Video recommendation method and server |
CN104572889A (en) * | 2014-12-24 | 2015-04-29 | 深圳市腾讯计算机***有限公司 | Method, device and system for recommending search terms |
CN104573109A (en) * | 2015-01-30 | 2015-04-29 | 深圳市中兴移动通信有限公司 | System, terminal and method for automatic recommendation based on group relation |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110941756A (en) * | 2018-09-25 | 2020-03-31 | 夏普株式会社 | Terminal device, information processing system, method, and recording medium |
CN109756563A (en) * | 2018-12-17 | 2019-05-14 | 平安科技(深圳)有限公司 | Information-pushing method, device, computer equipment and storage medium |
CN110929170A (en) * | 2019-11-22 | 2020-03-27 | 腾讯科技(深圳)有限公司 | Friend recommendation method, device, equipment and storage medium for social group |
CN111130992A (en) * | 2019-11-22 | 2020-05-08 | 北京达佳互联信息技术有限公司 | Group recommendation method and device, electronic equipment and storage medium |
CN110929170B (en) * | 2019-11-22 | 2024-06-28 | 腾讯科技(深圳)有限公司 | Friend recommendation method, device and equipment of social group and storage medium |
US11470032B2 (en) | 2019-11-22 | 2022-10-11 | Beijing Dajia Internet Information Technology Co., Ltd. | Method for recommending groups and related electronic device |
CN112989276A (en) * | 2019-12-12 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Evaluation method and device of information push system |
CN111523035B (en) * | 2020-04-24 | 2023-04-18 | 上海异势信息科技有限公司 | Recommendation method, device, server and medium for APP browsing content |
CN111523035A (en) * | 2020-04-24 | 2020-08-11 | 上海异势信息科技有限公司 | Recommendation method, device, server and medium for APP browsing content |
CN114765624A (en) * | 2020-12-31 | 2022-07-19 | 北京达佳互联信息技术有限公司 | Information recommendation method and device, server and storage medium |
CN112769676A (en) * | 2020-12-31 | 2021-05-07 | 上海掌门科技有限公司 | Method and equipment for providing information in group |
CN112769676B (en) * | 2020-12-31 | 2022-12-30 | 上海掌门科技有限公司 | Method and equipment for providing information in group |
CN114765624B (en) * | 2020-12-31 | 2024-04-30 | 北京达佳互联信息技术有限公司 | Information recommendation method, device, server and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108304428A (en) | Information recommendation method and device | |
US10394953B2 (en) | Meme detection in digital chatter analysis | |
US9152969B2 (en) | Recommendation ranking system with distrust | |
CN103324636B (en) | The system and method for commending friends in social networks | |
TW202007178A (en) | Method, device, apparatus, and storage medium of generating features of user | |
CN108369715B (en) | Interactive commentary based on video content characteristics | |
US8346746B2 (en) | Aggregation, organization and provision of professional and social information | |
US8615404B2 (en) | Self-describing data framework | |
US9436919B2 (en) | System and method of tuning item classification | |
US9230030B2 (en) | Method and system for mining websites | |
US20140074545A1 (en) | Human workflow aware recommendation engine | |
US20150286709A1 (en) | Method and system for retrieving information from knowledge-based assistive network to assist users intent | |
US9152709B2 (en) | Cross-domain topic space | |
CN108431852A (en) | Computer resource ranking for interconnected users profile | |
MX2012003721A (en) | Systems and methods for social graph data analytics to determine connectivity within a community. | |
Atzmueller et al. | Ubicon and its applications for ubiquitous social computing | |
US11444909B2 (en) | Latent user communities | |
US10795899B2 (en) | Data discovery solution for data curation | |
US10977677B2 (en) | Contact importer | |
CN108345601A (en) | Search result ordering method and device | |
US11036700B2 (en) | Automatic feature generation for machine learning in data-anomaly detection | |
CN107809370B (en) | User recommendation method and device | |
US20230037222A1 (en) | Method, apparatus and computer program product for generating tiered search index fields in a group-based communication platform | |
US10733244B2 (en) | Data retrieval system | |
US20160132771A1 (en) | Application Complexity Computation |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180720 |