CN103138954B - A kind of method for pushing of recommendation items, system and recommendation server - Google Patents
A kind of method for pushing of recommendation items, system and recommendation server Download PDFInfo
- Publication number
- CN103138954B CN103138954B CN201110397617.8A CN201110397617A CN103138954B CN 103138954 B CN103138954 B CN 103138954B CN 201110397617 A CN201110397617 A CN 201110397617A CN 103138954 B CN103138954 B CN 103138954B
- Authority
- CN
- China
- Prior art keywords
- user
- list
- recommended
- strategy
- active user
- 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.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of method for pushing of recommendation items, system and recommendation server, when determining the operation trigger recommendation event of active user, according to active user to each targeted customer in the Generalization bounds of current item to be recommended and the buddy list of active user to the reception strategy of current item to be recommended, produce candidate target user list; To determine according to candidate target user list and after the list of targeted subscribers returned, in list of targeted subscribers, each targeted customer recommends current item to be recommended receiving active user; For sending the active user of recommendation items, the targeted customer that can reduce to adopting probability low pushes recommendation items, has saved recommendation resource, and improving recommendation items by the basis of success rate of adopting, adds the flexibility that active user pushes; For pushed targeted customer, avoid and receive a large amount of uninterested recommendation items, saved recommendation resource, improve the probability adopted to current item to be recommended.
Description
Technical field
The present invention relates to business support field, particularly relate to a kind of method for pushing of recommendation items, system and recommendation server.
Background technology
Along with the rise of ecommerce, social media website, recommended technology has been widely applied in the application of many hot topics such as Taobao, bean cotyledon, Google's news, Amazon, popular comment net.Traditional commending system is mainly divided into two classes: content-based recommendation system and collaborative filtering system.In content-based recommendation system, input data are treated to user profiles one by one, each user profiles is generally expressed as a characteristic vector, recommended candidate information is carried out similar process, then carry out Similarity Measure with the profile of targeted customer, the candidate information closest to user profiles is pushed to user as recommendation items.In collaborative filtering system, user behavior data is used to calculate the similarity between user or between recommended candidate information, and recommendation results draws according to after the further weighting of this similitude.
Typical in the recommendation of social networks at one, the recommendation items that self likes can be sent to other user in social networks and receive the recommendation items pushed by other users from other user by user, but, due to the complexity that social networks is formed, the recommendation items that the good friend of certain user may push him is also lost interest in, simultaneously, certain user also can receive a large amount of self and uninterested recommendation items, cause the waste recommending resource, also reduce recommendation items by the success rate adopted; And after user receives recommendation items repeatedly from a good friend there, the possibility adopting recommendation items also can reduce gradually, makes overall recommendation items be reduced by the success rate adopted.
Summary of the invention
Embodiments provide a kind of method for pushing of recommendation items, system and recommendation server, cause recommend the waste of resource and recommendation items by the not high problem of the success rate adopted in order to solve existing method for pushing.
The method for pushing of a kind of recommendation items that the embodiment of the present invention provides, comprising:
When determining the operation trigger recommendation event of active user, obtain the buddy list of described active user according to the mark of described active user;
According to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
Described candidate target user list is presented to described active user, receives described active user and determine and the list of targeted subscribers returned according to described candidate target user list;
Send according to the confirmation that described active user sends the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended.
A kind of recommendation server that the embodiment of the present invention provides, comprising:
Acquisition module, for when determining the operation trigger recommendation event of active user, obtains the buddy list of described active user according to the mark of described active user;
Computing module, for according to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
Confirm module, for described candidate target user list is presented to described active user, receive described active user and determine and the list of targeted subscribers returned according to described candidate target user list;
Recommending module, send for the confirmation sent according to described active user the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended.
The supplying system of a kind of recommendation items that the embodiment of the present invention provides, comprising: recommendation server, service server and information promulgating platform server;
Described recommendation server, for when determining the operation trigger recommendation event of active user, obtains the buddy list of described active user according to the mark of described active user; According to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list; Described candidate target user list is presented to described active user, receives described active user and determine and the list of targeted subscribers returned according to described candidate target user list; Send according to the confirmation that described active user sends the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended;
Described service server, for providing friend information and the recommendation items information of user;
Described information promulgating platform server, the item current to be recommended for being pushed by described recommendation server is published to the receiving platform of targeted customer.
The beneficial effect of the embodiment of the present invention comprises:
The method for pushing of a kind of recommendation items that the embodiment of the present invention provides, system and recommendation server, when determining the operation trigger recommendation event of active user, according to active user to each targeted customer in the Generalization bounds of current item to be recommended and the buddy list of active user to the reception strategy of current item to be recommended, produce candidate target user list; To determine according to candidate target user list and after the list of targeted subscribers returned, in list of targeted subscribers, each targeted customer recommends current item to be recommended receiving active user.The method for pushing of the recommendation items that the embodiment of the present invention provides is when producing the list of targeted subscribers of current item to be recommended, the Generalization bounds of comprehensive reference between active user and targeted customer and receive strategy, filter out the targeted customer of applicable current item to be recommended, and the final each targeted customer confirmed via active user in list of targeted subscribers, for the active user sending recommendation items, the targeted customer that can reduce to adopting probability low pushes recommendation items, save recommendation resource, and improving recommendation items by the basis of success rate of adopting, add the flexibility that active user pushes, for pushed targeted customer, avoid and receive a large amount of uninterested recommendation items, saved recommendation resource, improve the probability adopted to current item to be recommended.
Accompanying drawing explanation
The flow chart of the method for pushing of the recommendation items that Fig. 1 provides for the embodiment of the present invention;
The flow chart of the acquisition Generalization bounds that Fig. 2 provides for the embodiment of the present invention;
The acquisition that Fig. 3 provides for the embodiment of the present invention receives tactful flow chart;
The structural representation of the recommendation server that Fig. 4 provides for the embodiment of the present invention;
The Organization Chart of the supplying system of the recommendation items of the example that Fig. 5 provides for the embodiment of the present invention;
The structural representation of the recommendation server in the example that Fig. 6 provides for the embodiment of the present invention.
Embodiment
Below in conjunction with accompanying drawing, the embodiment of the method for pushing of the recommendation items that the embodiment of the present invention provides, system and recommendation server is described in detail.
The method for pushing of a kind of recommendation items that the embodiment of the present invention provides, as shown in Figure 1, idiographic flow comprises the following steps:
S101, when determining the operation trigger recommendation event of active user, obtain the buddy list of active user according to the mark of active user;
S102, according to active user to each targeted customer in the Generalization bounds of current item to be recommended and buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
S103, candidate target user list is presented to active user, receive active user and determine and the list of targeted subscribers returned according to candidate target user list;
S104, the instruction recommended according to the confirmation transmission of active user's transmission, in list of targeted subscribers, each targeted customer pushes current item to be recommended.
Below the specific implementation of above steps is described in detail.
In above-mentioned steps S101, the operation of active user's trigger recommendation event can include but not limited to following several event: (1) active user completes the payment of online order; (2) active user completes the download to setting software; (3) active user completes and makes comments to set information; (4) active user selects Information Sharing to the friend of oneself.The meeting trigger recommendation event when active user completes above-mentioned event, system can obtain the buddy list of active user according to the mark of active user, in the specific implementation, system from the buddy list of the thin middle importing active user of social networks, outside social website, instant message software or e-mail address, can not limit the source of buddy list at this.
Preferably, in above-mentioned steps S102, the detailed process producing candidate target user list can comprise the following steps:
First, according to the Generalization bounds of active user to current item to be recommended, each targeted customer in buddy list is screened, obtain the candidate target user list meeting Generalization bounds;
Then, according to the reception strategy of targeted customer each in buddy list to current item to be recommended, the targeted customer met in the candidate target user list of Generalization bounds is screened, obtain meeting the candidate target user list receiving strategy.
The candidate target user list obtained by above-mentioned twice screening had both met the Generalization bounds of active user, meet again the reception strategy of targeted customer, like this, for the active user sending recommendation items, active user can be avoided recommendation items to be pushed to the uninterested targeted customer there of this recommendation items, save recommendation resource, for pushed targeted customer, avoid and receive a large amount of uninterested recommendation items, save recommendation resource, improve the probability adopted to current item to be recommended.
Particularly, the Generalization bounds of the user in the said method that provides of the embodiment of the present invention and the acquisition receiving strategy can be realized by following step:
For the Generalization bounds obtaining user, as shown in Figure 2, can comprise the following steps:
S201, for each user reached the standard grade, obtain the Generalization bounds that user is arranged each current item to be recommended;
S202, according to each targeted customer in the Generalization bounds of user and the buddy list of user to the reception strategy of current item to be recommended, prediction buddy list in targeted customer ratio is adopted to current item to be recommended; Such as: the recommendation of prediction user by how many good friends can be received, and by possibility that these good friends adopt;
S203, will predict the outcome and present to user;
S204, when receive user send Generalization bounds confirm instruction time, preserve user arrange Generalization bounds;
S205, when receive user send Generalization bounds upgrade instruction time, preserve user upgrade Generalization bounds.
Similarly, for the acquisition reception strategy of user and the Generalization bounds similar process of above-mentioned acquisition user, as shown in Figure 3, can be realized by following step:
S301, for each user reached the standard grade, obtain the reception strategy that user is arranged each current item to be recommended;
S302, to the Generalization bounds of current item to be recommended, predict the recommended amount that this user can receive over a period to come according to each targeted customer in the reception strategy of user and the buddy list of user;
S303, will predict the outcome and present to user;
S304, when receive user send reception strategy confirm instruction time, preserve user arrange reception strategy;
S305, when receive user send the instruction of reception policy update time, preserve user upgrade reception strategy.
In the specific implementation, the step S301 ~ S305 obtaining the step S201 ~ S205 of the Generalization bounds of user and the reception strategy of acquisition user can carry out also can carrying out respectively simultaneously, does not limit at this.
Preferably, in the step S103 of the said method that the embodiment of the present invention provides, to determine according to candidate target user list and before the list of targeted subscribers returned reception active user, when getting active user and revising its Generalization bounds, candidate target user list will be upgraded, and the candidate target user list after upgrading is presented to active user.
Particularly, in the said method that the embodiment of the present invention provides, Generalization bounds or the reception strategy of use can be one of following strategy:
The combined strategy of black and white lists strategy, circle of influence strategy, preference strategy and circle of influence strategy and preference strategy; Wherein, circle of influence strategy be according to active user to the recommendation number of times of each targeted customer in its buddy list and each targeted customer the recommendation to active user adopt situation arrange strategy.
Below each strategy is described in detail.
Black and white lists strategy: buddy list is divided into two lists, all good friends in white list are as targeted customer, and all good friends in blacklist are by disallowable.
Circle of influence strategy: be made up of three basic substrategys, connect by logical "and" or logical "or" between substrategy, to reach different filter effects, particularly, substrategy is: 1) active user sends new recommendation items to the good friend adopting its recommendation items within the time interval; 2) good friend that active user recommends number of times to be less than k to transmission accumulative within the time interval sends new recommendation items; 3) active user sends new recommendation items to receiving its n recommendation items recently and adopting one of them individual good friend.
Such as: assuming that only adopt 1) or 2) strategy, the time interval is set as 1 week, the accumulative recommendation number of times upper limit k sent is set as 7, so, if user A not yet reaches 7 to the recommendation transmission times of good friend B in one week, or exceeded 7 but good friend B once adopted a certain bar wherein, so good friend B as targeted customer, otherwise will remove by good friend B from targeted customer.
Utilize circle of influence strategy, active user can be avoided a large amount of items to be recommended to be sent to those and recommend to continue the good friend that do not pay close attention to there to it.Meanwhile, can also realize by sending the degree of concern soundd out good friend and recommend it on a small quantity.
Preference strategy: by arranging a threshold value, being greater than the good friend of threshold value as targeted customer using to the preference of current item to be recommended, because the computational methods of preference belong to prior art, not elaborating at this.
In the particular embodiment, the preference of good friend can be normalized according to certain principle that (such as linear normalization is to [0 by the arranging of threshold value, 1] interval), using good friend for average preference's degree of current item to be recommended as threshold value, or a concrete numerical value is set as required, such as, with 0 for threshold value can show the intimate filtration of preference out by all to current item to be recommended.
Preferably, the confirmation that step S104 in the method that the embodiment of the present invention provides sends according to active user sends the instruction recommended, in list of targeted subscribers, each targeted customer pushes current item to be recommended, in the specific implementation, a unique mark can be added in the access of the item current to be recommended generated connects, the recommendation items that can be identified user's click by this mark is which user concrete sends, like this, by recording user for the click of recommendation results and purchase conversion situation, in order to providing reference to subsequent user Provisioning Policy.
Based on same inventive concept, the embodiment of the present invention additionally provides the supplying system of a kind of recommendation server and recommendation items, the principle of dealing with problems due to this server and system is similar to the method for pushing of aforementioned a kind of recommendation items, therefore the enforcement of this server and system see the enforcement of method, can repeat part and repeats no more.
A kind of recommendation server that the embodiment of the present invention provides, as shown in Figure 4, comprising:
Acquisition module 401, for when determining the operation trigger recommendation event of active user, obtains the buddy list of active user according to the mark of active user;
Computing module 402, for according to active user to each targeted customer in the Generalization bounds of current item to be recommended and buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
Confirm module 403, for candidate target user list is presented to active user, receive active user and determine and the list of targeted subscribers returned according to candidate target user list;
Recommending module 404, send for the confirmation sent according to active user the instruction recommended, in list of targeted subscribers, each targeted customer pushes current item to be recommended.
Further, the above-mentioned server that the embodiment of the present invention provides, as shown in Figure 4, can also comprise: tactful preservation module 405, for for each user reached the standard grade, obtains the Generalization bounds of user to each current item setting to be recommended tactful with reception; According to each targeted customer in the Generalization bounds of user and the buddy list of user to the reception strategy of current item to be recommended, the targeted customer in prediction buddy list adopts ratio to current item to be recommended; Or according to the Generalization bounds of each targeted customer in the reception strategy of user and the buddy list of user, the recommended amount that prediction user can receive over a period to come; To predict the outcome and present to user; When receiving the Generalization bounds of user's transmission or receiving tactful confirmation instruction, preserve the Generalization bounds of user's setting or receive tactful, when the Generalization bounds that reception user sends or reception policy update instruction, preserving Generalization bounds or the reception strategy of user's renewal.
Further, computing module 402 in the above-mentioned server that the embodiment of the present invention provides, specifically for according to the Generalization bounds of active user to current item to be recommended, each targeted customer in buddy list is screened, obtain the candidate target user list meeting Generalization bounds; According to the reception strategy of targeted customer each in buddy list to current item to be recommended, the targeted customer met in the candidate target user list of Generalization bounds is screened, obtain meeting the candidate target user list receiving strategy.
Further, confirmation module 403 in the above-mentioned server that the embodiment of the present invention provides, also for confirm according to candidate target user list reception active user and before the list of targeted subscribers returned, when getting active user and revising Generalization bounds, upgrade candidate target user list, and the candidate target user list after upgrading is presented to active user.
The embodiment of the present invention additionally provides a kind of supplying system of recommendation items, comprising: recommendation server, service server and information promulgating platform server;
Recommendation server, for when determining the operation trigger recommendation event of active user, obtains the buddy list of active user according to the mark of active user; According to active user to each targeted customer in the Generalization bounds of current item to be recommended and buddy list to the reception strategy of current item to be recommended, produce candidate target user list; Candidate target user list is presented to active user, receives active user and determine and the list of targeted subscribers returned according to candidate target user list; Send according to the confirmation that active user sends the instruction recommended, in list of targeted subscribers, each targeted customer pushes current item to be recommended;
Service server, for providing friend information and the recommendation items information of user;
Information promulgating platform server, the item current to be recommended for being pushed by recommendation server is published to the receiving platform of targeted customer.
Further, can also comprise in the said system that the embodiment of the present invention provides: external data source;
Recommendation server, also for obtaining the friend information of user from external data source.
Below by instantiation, the said system that the embodiment of the present invention provides is described, as shown in Figure 5, recommendation server, service server, information promulgating platform server and external data source (information promulgating platform server and external data source not shown) is comprised.
Wherein, the concrete structure of recommendation server as shown in Figure 6, can comprise with lower module: recommend trigger control module 601, recommendation target computing module 602, social networks administration module 603, preference computing module 604, Generalization bounds control module 605 and recommend sending module 606.Wherein, recommend trigger control module 601, social administration module 603, Generalization bounds control module 605, recommend sending module 606 with the service server (such as e-commerce platform) in system, external data source (good friend that such as social network sites is open obtains Web service), information promulgating platform server (such as microblogging distribution platform), information transmission may occur.
Particularly, recommending trigger control module 601, for receiving the recommendation trigger event from service server, and submitting to recommendation target computing module 602 request of propelling movement to.Recommend trigger event to be determined by the service needed of system, typical event includes but not limited to: 1) user completes the payment of online order; 2) user has downloaded certain software in application store; 3) user has delivered comment to certain information; 4) user initiatively selects Information Sharing to the friend of oneself.
Recommend target computing module 602, for receiving the propelling movement request from recommending trigger control module 601, this propelling movement request comprises the id of an active user id, current item to be recommended, recommend target computing module 602 can carry out alternately, determining the transmission targeted user population of applicable active user and current item to be recommended with social networks administration module 603, preference computing module 604 and Generalization bounds control module 605 according to these two marks.
Particularly, target computing module 602 is recommended to comprise: social networks request module 6021, strategy request module 6022, preference request module 6023, historical behavior request module 6024 and policy filtering module 6025.
When embody rule, target computing module 602 is recommended to call social networks request module 6021 and strategy request module 6022 successively, and whether call preference request module 6023 and historical behavior request module 6024 according to the strategy decision got, to obtain required user's buddy list, user's historical behavior and good friend are to information such as the preference of current item to be recommended, afterwards, target computing module 602 regulative strategy filtering module 6025 pairs of user's buddy lists are recommended to carry out policy filtering, produce final candidate target user list, be supplied to and recommend sending module 606 to carry out user's confirmation, and complete transmission.
Social networks administration module 603, for maintain the buddy list of user, after receiving the request recommending target computing module 602, returns its buddy list according to active user id.
Particularly, social networks administration module 603 comprises: inner good friend's administration module 6031 and outside good friend's administration module 6032, be respectively used to the good friend of leading subscriber in internal system and the good friend of its exterior, wherein, inner good friend refers to that other users directly add as a friend by the friend making function that user utilizes internal system to provide; Outside good friend refers to the friend relation that user imports from the data sources such as mail address book, social network sites and chat tool.When embody rule, outside good friend's administration module 6032 also may comprise an invitation of inviting external user to become internal user and confirm function, or, also can utilize the API that the such as open platform such as Google OpenSocial, Facebook Friend Connect provides, direct and external data source carries out alternately.
Preference computing module 604, for calculating the fancy grade of targeted customer to current recommendation items, this module receives the buddy list of user and the id of current item to be recommended from recommending target computing module 602, then to return in buddy list all targeted customers to the preference of current recommendation items, the data of preference are from the user behavior (such as buy, click, download, program request etc.) of operation system record.
Particularly, preference computing module 604 comprises: content-based preference computing module 6041, the preference computing module 6042 based on collaborative filtering and the preference computing module 6043 based on mixed strategy.The text feature of content-based preference computing module 6041 according to the history access object of user and the text feature of item to be recommended, calculate the preference that this user treats recommendation items.Generally speaking, the algorithm that this module adopts all needs the preference of user and item to be recommended to be expressed as property vector.Preference computing module 6042 based on collaborative filtering passes through the similitude of measure user history access record, according to the Visitor Logs of other users similar to user, calculates the preference of this user to current item to be recommended.Preference computing module 6043 based on mixed strategy combines the feature of content-based preference computing module 6041 and the preference computing module 6042 based on collaborative filtering, to promote the comprehensive of preference calculating.In the specific implementation, the preference request (comprising mark and the buddy list of current item to be recommended) that can send according to preference request module 6023, calculate targeted customer in buddy list to the preference of current item to be recommended, and return to buddy list and recommend target computing module 602 to calculate for follow-up policy filtering.
Generalization bounds control module 605, arranges corresponding Generalization bounds and reception strategy for user according to the needs of oneself.These strategies include but are not limited to circle of influence strategy, black and white lists strategy, preference strategy etc.
Particularly, Generalization bounds control module 605 comprises: black and white lists policy module 6051, circle of influence policy module 6052 and preference policy module 6053, and user can by the Generalization bounds of above-mentioned three module installation three types: black and white lists strategy, circle of influence strategy, preference strategy.
Recommend sending module 606, for according to the output recommending target computing module 602, recommendation items is sent to corresponding targeted user population.This module, before transmission recommendation items, comprises one and is edited and the process confirmed by active user.
Particularly, sending module 606 is recommended to comprise: inner sending module 6061 and outside sending module 6062.Inner sending module 6061, for recommendation items is presented in intrasystem displayed page, when after recommended targeted customer's login system, will see the content of recommendation items in respective page.The interface of outside sending module 6062 for utilizing external system to provide, recommendation items is sent to the displaying position of external system, wherein, Typical external system approach comprises the Web service interface etc. that Email, microblogging and social network sites provide.Further, in order to track user accepts situation for recommendation items, recommend sending module 606 to add a unique identification by the access links generated, that can identify that user x clicks by this mark is recommendation items instead of the user z of user y.These data to the click/purchase conversion situation of recommendation items, and are supplied to native system use by the interface of Generalization bounds control module 605 by operation system physical record user.Finally, for the needs of respecting privacy of user, in any embodiment, recommend sending module that user all should be provided to confirm function, and allow user manually to adjust final list of targeted subscribers.
The supplying system of the above-mentioned recommendation items that the embodiment of the present invention provides just illustrates, the supplying system of recommendation items also comprises other forms in the specific implementation, does not elaborate at this.
Through the above description of the embodiments, those skilled in the art can be well understood to the embodiment of the present invention can by hardware implementing, and the mode that also can add necessary general hardware platform by software realizes.Based on such understanding, the technical scheme of the embodiment of the present invention can embody with the form of software product, it (can be CD-ROM that this software product can be stored in a non-volatile memory medium, USB flash disk, portable hard drive etc.) in, comprise some instructions and perform method described in each embodiment of the present invention in order to make a computer equipment (can be personal computer, server, or the network equipment etc.).
It will be appreciated by those skilled in the art that accompanying drawing is the schematic diagram of a preferred embodiment, the module in accompanying drawing or flow process might not be that enforcement the present invention is necessary.
It will be appreciated by those skilled in the art that the module in the device in embodiment can carry out being distributed in the device of embodiment according to embodiment description, also can carry out respective change and be arranged in the one or more devices being different from the present embodiment.The module of above-described embodiment can merge into a module, also can split into multiple submodule further.
The invention described above embodiment sequence number, just to describing, does not represent the quality of embodiment.
The method for pushing of a kind of recommendation items that the embodiment of the present invention provides, system and recommendation server, when determining the operation trigger recommendation event of active user, according to active user to each targeted customer in the Generalization bounds of current item to be recommended and the buddy list of active user to the reception strategy of current item to be recommended, produce candidate target user list; To determine according to candidate target user list and after the list of targeted subscribers returned, in list of targeted subscribers, each targeted customer recommends current item to be recommended receiving active user.The method for pushing of the recommendation items that the embodiment of the present invention provides is when producing the list of targeted subscribers of current item to be recommended, the Generalization bounds of comprehensive reference between active user and targeted customer and receive strategy, filter out the targeted customer of applicable current item to be recommended, and the final each targeted customer confirmed via active user in list of targeted subscribers, for the active user sending recommendation items, the targeted customer that can reduce to adopting probability low pushes recommendation items, save recommendation resource, and improving recommendation items by the basis of success rate of adopting, add the flexibility that active user pushes, for pushed targeted customer, avoid and receive a large amount of uninterested recommendation items, saved recommendation resource, improve the probability adopted to current item to be recommended.
Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.
Claims (9)
1. a method for pushing for recommendation items, is characterized in that, comprising:
When determining the operation trigger recommendation event of active user, obtain the buddy list of described active user according to the mark of described active user;
According to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
Described candidate target user list is presented to described active user, receives described active user and determine and the list of targeted subscribers returned according to described candidate target user list;
Send according to the confirmation that described active user sends the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended; Wherein before determining the operation trigger recommendation event of active user, also comprise: for each user reached the standard grade, obtain user to the Generalization bounds of each current item setting to be recommended and reception strategy; According to each targeted customer in the Generalization bounds of described user and the buddy list of described user to the reception strategy of current item to be recommended, predict that the targeted customer in described buddy list adopts ratio to current item to be recommended; Or according to the Generalization bounds of each targeted customer in the reception strategy of described user and the buddy list of described user, predict the recommended amount that described user can receive over a period to come; To predict the outcome and present to described user; When receiving the Generalization bounds of described user transmission or receiving strategy confirmation instruction, preserve the Generalization bounds of described user setting or receive strategy, when receiving the Generalization bounds of described user transmission or receiving policy update instruction, preserve the Generalization bounds of described user renewal or receive strategy.
2. the method for claim 1, is characterized in that, according to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list, specifically comprise:
According to the Generalization bounds of described active user to current item to be recommended, each targeted customer in described buddy list is screened, obtain the candidate target user list meeting described Generalization bounds;
According to the reception strategy of targeted customer each in described buddy list to described current item to be recommended, the targeted customer met in the candidate target user list of described Generalization bounds is screened, obtain meeting the described candidate target user list receiving strategy.
3. method as claimed in claim 2, it is characterized in that, described Generalization bounds or described reception strategy are one of following strategy:
The combined strategy of black and white lists strategy, circle of influence strategy, preference strategy and circle of influence strategy and preference strategy; Wherein, described circle of influence strategy be according to described active user to the recommendation number of times of each targeted customer in its buddy list and each targeted customer the recommendation to described active user adopt situation arrange strategy.
4. the method as described in any one of claim 1-3, is characterized in that, in reception, described active user to determine according to described candidate target user list and before the list of targeted subscribers returned, also to comprise:
When getting described active user and revising described Generalization bounds, upgrade candidate target user list, and the candidate target user list after upgrading is presented to described active user.
5. a recommendation server, is characterized in that, comprising:
Acquisition module, for when determining the operation trigger recommendation event of active user, obtains the buddy list of described active user according to the mark of described active user;
Computing module, for according to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list;
Confirm module, for described candidate target user list is presented to described active user, receive described active user and determine and the list of targeted subscribers returned according to described candidate target user list;
Recommending module, send for the confirmation sent according to described active user the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended;
Tactful preservation module, for for each user reached the standard grade, obtains the Generalization bounds of user to each current item setting to be recommended tactful with reception; According to each targeted customer in the Generalization bounds of described user and the buddy list of described user to the reception strategy of current item to be recommended, predict that the targeted customer in described buddy list adopts ratio to current item to be recommended; Or according to the Generalization bounds of each targeted customer in the reception strategy of described user and the buddy list of described user, predict the recommended amount that described user can receive over a period to come; To predict the outcome and present to described user; When receiving the Generalization bounds of described user transmission or receiving strategy confirmation instruction, preserve the Generalization bounds of described user setting or receive strategy, when receiving the Generalization bounds of described user transmission or receiving policy update instruction, preserve the Generalization bounds of described user renewal or receive strategy.
6. server as claimed in claim 5, it is characterized in that, described computing module, specifically for according to the Generalization bounds of described active user to current item to be recommended, each targeted customer in described buddy list is screened, obtains the candidate target user list meeting described Generalization bounds; According to the reception strategy of targeted customer each in described buddy list to described current item to be recommended, the targeted customer met in the candidate target user list of described Generalization bounds is screened, obtain meeting the described candidate target user list receiving strategy.
7. the server as described in claim 5 or 6, it is characterized in that, described confirmation module, also for confirm according to described candidate target user list the described active user of reception and before the list of targeted subscribers returned, when getting described active user and revising described Generalization bounds, upgrade candidate target user list, and the candidate target user list after upgrading is presented to described active user.
8. a supplying system for recommendation items, is characterized in that, comprising: recommendation server, service server and information promulgating platform server;
Described recommendation server, for when determining the operation trigger recommendation event of active user, obtains the buddy list of described active user according to the mark of described active user; According to described active user to each targeted customer in the Generalization bounds of current item to be recommended and described buddy list to the reception strategy of current item to be recommended, produce candidate target user list; Described candidate target user list is presented to described active user, receives described active user and determine and the list of targeted subscribers returned according to described candidate target user list; Send according to the confirmation that described active user sends the instruction recommended, in described list of targeted subscribers, each targeted customer pushes described current item to be recommended;
Wherein, described recommendation server before determining the operation trigger recommendation event of active user, also for: for each user reached the standard grade, obtain Generalization bounds that user arranges each current item to be recommended and receive strategy; According to each targeted customer in the Generalization bounds of described user and the buddy list of described user to the reception strategy of current item to be recommended, predict that the targeted customer in described buddy list adopts ratio to current item to be recommended; Or according to the Generalization bounds of each targeted customer in the reception strategy of described user and the buddy list of described user, predict the recommended amount that described user can receive over a period to come; To predict the outcome and present to described user; When receiving the Generalization bounds of described user transmission or receiving strategy confirmation instruction, preserve the Generalization bounds of described user setting or receive strategy, when receiving the Generalization bounds of described user transmission or receiving policy update instruction, preserve the Generalization bounds of described user renewal or receive strategy;
Described service server, for providing friend information and the recommendation items information of user;
Described information promulgating platform server, the item current to be recommended for being pushed by described recommendation server is published to the receiving platform of targeted customer.
9. system as claimed in claim 8, is characterized in that, also comprise: external data source;
Described recommendation server, also for obtaining the friend information of user from described external data source.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110397617.8A CN103138954B (en) | 2011-12-02 | 2011-12-02 | A kind of method for pushing of recommendation items, system and recommendation server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110397617.8A CN103138954B (en) | 2011-12-02 | 2011-12-02 | A kind of method for pushing of recommendation items, system and recommendation server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103138954A CN103138954A (en) | 2013-06-05 |
| CN103138954B true CN103138954B (en) | 2015-11-04 |
Family
ID=48498304
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110397617.8A Active CN103138954B (en) | 2011-12-02 | 2011-12-02 | A kind of method for pushing of recommendation items, system and recommendation server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103138954B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103312816B (en) * | 2013-07-01 | 2016-04-27 | 青岛海信传媒网络技术有限公司 | A kind of message distributing method and equipment |
| CN104281599A (en) * | 2013-07-02 | 2015-01-14 | 北京千橡网景科技发展有限公司 | Method and device for recommending information to user in social network |
| CN104144196A (en) * | 2013-07-05 | 2014-11-12 | 腾讯科技（深圳）有限公司 | Webpage information sharing method, device and server |
| CN103561067A (en) * | 2013-10-24 | 2014-02-05 | 青岛英网资讯股份有限公司 | Invitation method based on cellphone applications |
| CN103685502B (en) * | 2013-12-09 | 2017-07-25 | 腾讯科技（深圳）有限公司 | A kind of information push method, apparatus and system |
| CN104765738B (en) * | 2014-01-03 | 2019-03-26 | 腾讯科技（北京）有限公司 | The method and system that the counter of user-generated content is controlled |
| CN105786813A (en) * | 2014-12-17 | 2016-07-20 | 北京邮电大学 | Method and device for sending task information |
| CN105893019A (en) * | 2015-12-15 | 2016-08-24 | 乐视移动智能信息技术（北京）有限公司 | User manual calling-out method and apparatus for smartphone |
| CN105677881B (en) * | 2016-01-12 | 2020-07-03 | 腾讯科技（深圳）有限公司 | Information recommendation method and device and server |
| CN107292646A (en) * | 2016-04-01 | 2017-10-24 | 中国移动通信集团河北有限公司 | A kind of method and apparatus of generation business |
| CN107979624B (en) * | 2016-10-24 | 2020-12-15 | 腾讯科技（深圳）有限公司 | Information pushing method and device and client with quick access function |
| CN106789564B (en) * | 2016-12-08 | 2019-12-20 | 广州爱九游信息技术有限公司 | Social invitation system, equipment, device and method |
| CN107103490B (en) * | 2017-04-11 | 2021-07-09 | 华为机器有限公司 | Data processing method, network server and terminal |
| CN107193984A (en) * | 2017-05-25 | 2017-09-22 | 上海喆之信息科技有限公司 | A kind of high-quality user's commending system |
| CN107249023B (en) * | 2017-05-27 | 2020-06-30 | 维沃移动通信有限公司 | Processing method and device for file sending |
| CN107766441B (en) * | 2017-09-21 | 2021-11-23 | 北京百度网讯科技有限公司 | Resource update reminding method and device based on artificial intelligence and storage medium |
| CN108038789A (en) * | 2017-11-01 | 2018-05-15 | 平安科技（深圳）有限公司 | Item recommendation method, electronic equipment and computer-readable recording medium |
| CN107835120A (en) * | 2017-11-03 | 2018-03-23 | 上海掌门科技有限公司 | For providing the method and apparatus to release news of target good friend |
| CN108737535B (en) * | 2018-05-14 | 2021-10-29 | 平安科技（深圳）有限公司 | Message pushing method, storage medium and server |
| CN111988642B (en) * | 2019-05-24 | 2022-05-24 | 北京达佳互联信息技术有限公司 | Method, device, server and storage medium for recommending videos |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101540739A (en) * | 2009-04-14 | 2009-09-23 | 腾讯科技（深圳）有限公司 | User recommendation method and user recommendation system |
| CN101916286A (en) * | 2010-08-23 | 2010-12-15 | 宇龙计算机通信科技(深圳)有限公司 | Information recommendation method and system |
| CN102118324A (en) * | 2010-01-06 | 2011-07-06 | 腾讯科技（深圳）有限公司 | Method and system for inviting friend to access related product |
| CN102129639A (en) * | 2011-01-28 | 2011-07-20 | Tcl集团股份有限公司 | Method and system for sharing application procedure |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8234193B2 (en) * | 2008-03-03 | 2012-07-31 | Wildfire Interactive, Inc. | Method and system for providing online promotions through a social network-based platform |
-
2011
- 2011-12-02 CN CN201110397617.8A patent/CN103138954B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101540739A (en) * | 2009-04-14 | 2009-09-23 | 腾讯科技（深圳）有限公司 | User recommendation method and user recommendation system |
| CN102118324A (en) * | 2010-01-06 | 2011-07-06 | 腾讯科技（深圳）有限公司 | Method and system for inviting friend to access related product |
| CN101916286A (en) * | 2010-08-23 | 2010-12-15 | 宇龙计算机通信科技(深圳)有限公司 | Information recommendation method and system |
| CN102129639A (en) * | 2011-01-28 | 2011-07-20 | Tcl集团股份有限公司 | Method and system for sharing application procedure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103138954A (en) | 2013-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103138954B (en) | A kind of method for pushing of recommendation items, system and recommendation server | |
| US10050879B2 (en) | Techniques for selecting content based on network conditions | |
| US9641470B2 (en) | Channel based communication and transaction system | |
| AU2014209386B2 (en) | Method and apparatus for building a user profile for personalization using interaction data | |
| RU2451986C2 (en) | Systems, apparatus and methods of creating recommendations | |
| US20120316921A1 (en) | Survey Administration System and Methods | |
| US20110055017A1 (en) | System and method for semantic based advertising on social networking platforms | |
| CN107820695A (en) | The technology automatically determined for form response | |
| US20160292204A1 (en) | System and method for compiling and dynamically updating a collection of frequently asked questions | |
| CN106789598B (en) | Social relation chain-based public number message pushing method, device and system | |
| US20170090690A1 (en) | Continuity of experience card for index | |
| WO2010021835A1 (en) | Determining user affinity towards applications on a social networking website | |
| CN104704522A (en) | Recommending native applications | |
| CN102567479A (en) | Aggregated profile and online concierge | |
| WO2014062762A1 (en) | Propagating information through networks | |
| AU2014374421B2 (en) | Object recommendation based upon similarity distances | |
| CN104919481A (en) | Spotting trends by identifying influential consumers | |
| WO2014139372A1 (en) | System and method for selecting and rendering content | |
| US10909569B1 (en) | Obtaining a composite prediction indicating a likelihood that a user of an online system will perform a conversion associated with a content item via one or more paths of events | |
| CN103400292A (en) | Message advertising method and device | |
| CN107370659A (en) | Health control public number extension system and method based on wechat group | |
| US10687105B1 (en) | Weighted expansion of a custom audience by an online system | |
| CN112446736A (en) | Click through rate CTR prediction method and device | |
| CN109033190A (en) | A kind of method for pushing of recommendation information, device and equipment | |
| US11188846B1 (en) | Determining a sequential order of types of events based on user actions associated with a third party system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |