CN103034508B - Software recommendation method and system - Google Patents
Software recommendation method and system Download PDFInfo
- Publication number
- CN103034508B CN103034508B CN201110304854.5A CN201110304854A CN103034508B CN 103034508 B CN103034508 B CN 103034508B CN 201110304854 A CN201110304854 A CN 201110304854A CN 103034508 B CN103034508 B CN 103034508B
- Authority
- CN
- China
- Prior art keywords
- customer group
- user
- software
- feature
- characteristic information
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of software recommendation method, comprises the following steps: server creates customer group feature database, stores customer group classification and corresponding customer group differentiation feature in described customer group feature database; Client extracts the characteristic information of user, and the characteristic information of described user is submitted to described server; With the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated by described server, determine the customer group classification belonging to described user; Described server recommends software according to the described customer group classification determined to described user.Above-mentioned software recommendation method, according to the software that the characteristic information of user recommends its owning user realm not corresponding to user, can improve the possibility that user accepts to recommend software, thus improve the success ratio of software recommend.In addition, a kind of software recommend system is also provided.
Description
[technical field]
The present invention relates to field of computer technology, particularly a kind of software recommendation method and system.
[background technology]
Along with the development of computer technology, client software is more and more extensive, throughout each classes such as instant messaging, audio and video playing, resource downloading, web page browsing, input method, system supplymentaries.The new software of client emerges in an endless stream, and on the one hand software business man wishes new software to be pushed to user, and user also wishes to touch the new software oneself liked on the other hand, to produce the demand of software recommend.
Traditional software recommend mode recommends software similar or similar with mounted software to user.Such as, user has downloaded a player software, and system can recommend another player software again.And in fact, for this kind of software of player, user often only needs installation a.So this software recommend has little significance to user, do not provide required for it or interested new software to user, thus user accept to recommend the success ratio of software also not high.
[summary of the invention]
Based on this, be necessary to provide a kind of software recommendation method that can improve software recommend success ratio.
A kind of software recommendation method, comprises the following steps:
Server creates customer group feature database, stores customer group classification and corresponding customer group differentiation feature in described customer group feature database;
Client extracts the characteristic information of user, and the characteristic information of described user is submitted to described server;
With the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated by described server, determine the customer group classification belonging to described user;
Described server recommends software according to the described customer group classification determined to described user.
Preferably, the step of described server establishment customer group feature database comprises:
Choose sample of users;
Described sample of users is classified, obtains the initial category of described sample of users;
Obtain the characteristic information of described sample of users;
Characteristic information according to described initial category and user creates customer group feature database.
Preferably, the step of the characteristic information of described acquisition sample of users is:
Obtain the user profile of sample of users, extract characteristic information relevant to software in described user profile;
The step that the described characteristic information according to described initial category and user creates customer group feature database is:
According to described initial category, cluster analysis is carried out to the characteristic information of described user;
Obtain customer group classification according to cluster analysis result, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature;
Described customer group classification, corresponding customer group are differentiated that feature and described customer group differentiate that the reference value of feature is stored in customer group feature database.
Preferably, with the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated by described server, determines that the step of the customer group classification belonging to described user is:
Obtain the eigenwert in the characteristic information of described user;
Calculate the distance between reference value that the described eigenwert described customer group not corresponding with all types of user realm differentiate feature;
Obtain the customer group classification that minimum described distance is corresponding, the customer group classification belonging to described user.
Preferably, described method also comprises:
The software recommend list that described Servers installed is corresponding with described customer group classification;
Described server recommends the step of software to be according to the described customer group classification determined to described user:
Corresponding software recommend list is chosen according to the described customer group classification determined;
Filter out from described software recommend list described user uninstalled or than the software version update of user installation software recommend give described user.
In addition, there is a need to provide a kind of software recommend system that can improve software recommend success ratio.
A kind of software recommend system, comprise server and carry out the client of data interaction with described server, described server comprises:
Creation module, for creating customer group feature database, stores customer group classification and corresponding customer group differentiation feature in described customer group feature database;
Described client comprises: information extraction modules, for extracting the characteristic information of user, and the characteristic information of described user is submitted to described server;
Described server also comprises:
Matching module, for the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated, determines the customer group classification belonging to described user;
Recommending module, for recommending software according to the described customer group classification determined to described user.
Preferably, described creation module comprises:
Module chosen by sample, for choosing sample of users;
Preliminary classification module, for classifying to described sample of users, obtains the initial category of described sample of users;
Characteristic information acquisition module, for obtaining the characteristic information of described sample of users;
Feature database creates submodule, creates customer group feature database for the characteristic information according to described initial category and user.
Preferably, described characteristic information acquisition module also for obtaining the user profile of described sample of users, extracts characteristic information relevant to software in described user profile;
Described feature database creates submodule and comprises:
Cluster Analysis module, for carrying out cluster analysis according to described initial category to the characteristic information of described user;
Customer group classification acquisition module, for obtaining customer group classification according to cluster analysis result, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature;
Customer group feature database, for storing described customer group classification, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature.Preferably, described matching module comprises:
Characteristic value acquisition module, for obtaining the eigenwert in the characteristic information of described user;
Distance calculation module, the distance between the reference value differentiating feature for calculating described eigenwert and all kinds of described customer group;
Class of subscriber determination module, for obtaining customer group classification corresponding to minimum described distance, the customer group classification belonging to described user.
Preferably, described server also comprises:
Module is set, for arranging the software recommend list corresponding with described customer group classification;
Described recommending module comprises:
Module is chosen in list, for choosing corresponding software recommend list according to the described customer group classification determined;
Screening module, for filter out from described software recommend list described user uninstalled or than the software version update of user installation software recommend give described user.
Above-mentioned software recommendation method and system, differentiate that feature is mated by the characteristic information of user client extracted with the customer group in the customer group feature database in server, determine the customer group classification belonging to user, right rear line recommends the software corresponding with customer group classification.Software interested to different customer group classifications is different, according to the software that the characteristic information of user recommends its owning user realm not corresponding to user, can improve the possibility that user accepts to recommend software, thus improve the success ratio of software recommend.
[accompanying drawing explanation]
Fig. 1 is the schematic flow sheet of the software recommendation method in an embodiment;
Fig. 2 is the schematic flow sheet of the establishment customer group feature database in an embodiment;
Fig. 3 is the schematic flow sheet of the establishment customer group feature database in another embodiment;
Fig. 4 is other schematic flow sheet of determination user owning user realm in an embodiment;
Fig. 5 is the schematic flow sheet recommending software in an embodiment according to the customer group classification determined to user;
Fig. 6 is the structural representation of the software recommend system in an embodiment;
Fig. 7 is the structural representation of the creation module in an embodiment;
Fig. 8 is the structural representation of the feature database establishment submodule in an embodiment;
Fig. 9 is the structural representation of the matching module in an embodiment;
Figure 10 is the structural representation of the software recommend system in another embodiment.
[embodiment]
As shown in Figure 1, in one embodiment, a kind of software recommendation method, comprises the following steps:
Step S10, server creates customer group feature database, stores customer group classification and corresponding customer group differentiation feature in customer group feature database.
Customer group classification carries out the rear plurality of classes obtained of classifying, as game user class, User class etc. to a large amount of users.Customer group differentiates that feature refers to the feature for differentiating the customer group classification belonging to user, such as, the differentiation feature of User class can be install the number of online class software, the number of the amusement class software of installation, the study class software of installation number, use time period etc. of software.
Step S20, client extracts the characteristic information of user, and the characteristic information of user is submitted to server.
Use user in the process of various software, the raw information of client records user, such as, filled software matrix, the set-up time of software, the frequency of utilization of software, the upgrading frequency of software, user browse webpage record etc.Client is characteristic information extraction from the raw information of record, and the characteristic information extracted can be used for determining that user's owning user realm is other.Concrete, characteristic information comprises all kinds of software number, software set-up time, all kinds of software application frequency, all kinds of software upgrading frequency, all kinds of web page browsing durations etc., and wherein, the classification of software can be online class, amusement class, game class, study class etc.
In one embodiment, can be extracted from the raw information of user for determining other characteristic information of user's owning user realm by client, and the characteristic information of extraction is submitted to server.In another embodiment, the raw information of user also directly can be submitted to server by client, by server characteristic information extraction from the raw information of user.
Step S30, the characteristic information of user and the customer group in customer group feature database are differentiated that feature is mated by server, determine the customer group classification belonging to user.
Concrete, server obtains and differentiates feature with the customer group that user's characteristic information is mated most, then the customer group classification that this differentiation feature is corresponding is the customer group classification belonging to user.
Step S40, server recommends software according to the customer group classification determined to user.
Different customer group classifications can be arranged recommends different software, and such as, for game user class, then recommended games class software, for office users class, then recommends office class software.
Above-mentioned software recommendation method, differentiate that feature is mated by the characteristic information of user client extracted with the customer group in the customer group feature database in server, determine the customer group classification belonging to user, right rear line recommends the software corresponding with customer group classification.Software interested to different customer group classifications is different, according to the software that the characteristic information of user recommends its owning user realm not corresponding to user, can improve the possibility that user accepts to recommend software, thus improve the success ratio of software recommend.
As shown in Figure 2, in one embodiment, the detailed process of server establishment customer group feature database is:
Step S110, chooses sample of users.
Server from store user data selected part user as sample of users, the sample of users chosen relates to different classifications as far as possible, as network bar users class, User class, child user class, the elderly's user class, software development user class, game user class etc.
Step S120, classifies to sample of users, obtains the initial category of sample of users.
Step S130, obtains the characteristic information of sample of users.
The user data stored in server comprises the raw information of user, therefore the characteristic information of user be need extract from raw information, all kinds of software number, software set-up time, all kinds of software application frequency, all kinds of software upgrading frequency, all kinds of web page browsing durations etc. comprised.
Step S140, the characteristic information according to initial category and user creates customer group feature database.
As shown in Figure 3, in another embodiment, the detailed process of server establishment customer group feature database is:
Step S101, chooses sample of users.
Step S102, classifies to sample of users, obtains the initial category of sample of users.
Step S103, obtains the user profile of sample of users, extracts characteristic information relevant to software in user profile.
User profile is the raw information of user, the characteristic information relevant to software is extracted from the raw information of user, concrete, by traditional Euclidean distance tolerance, by probability metrics criterion, with divergence criterion function, to combine based on one or more in the methods such as differentiation entropy minimization and extract the characteristic information relevant to software.The characteristic information relevant to software such as extracted comprise the number of all kinds of softwares of installation, the set-up time of all kinds of software, the frequency of utilization of all kinds of software, the upgrading frequency of all kinds of software and all kinds of webpage browse duration etc.
Step S104, carries out cluster analysis according to initial category to the characteristic information of user.
Concrete, traditional C mean algorithm, ISODATA algorithm can be adopted, combine based on any one or a few method in the method such as similarity measurements quantity algorithm, adjacent function criterion algorithm of sample and core and cluster analysis is carried out to the characteristic information of user.
Step S105, obtains customer group classification according to cluster analysis result, corresponding customer group differentiates feature and the reference value of customer group differentiation feature.
If for the some features in characteristic information, the corresponding eigenwert of all sample of users of a certain initial category all relatively, and differ greatly with the corresponding eigenwert of the sample of users of other initial category, illustrate that this feature not only has cohesion to this customer group, also this customer group and other customer group are separated by this feature, then this is characterized as the differentiation feature of this customer group.Contrary, if for the sample of users of some initial category, there is very big-difference in the eigenwert of the some features in its characteristic information, then as this initial user realm, other does not differentiate feature to this feature.Such as in 100 features of game user class, there is very large difference in the eigenwert of this feature of installation number of tool-class software, then this feature of installation number of tool-class software is not as the differentiation feature of game user class.
Further, the reference value of this differentiation feature can be determined according to the corresponding eigenwert of all sample of users of this initial category.Mode in the concrete mean value can getting all eigenwerts corresponding to this differentiation feature or all eigenwerts is as the reference value of this differentiation feature of this customer group.
Further, determine final customer group classification, inappropriate preliminary classification is deleted or merges.If for each feature in characteristic information, between the sample of users eigenwert of a certain initial category, all there is very large difference, illustrate that the division of this initial category is nonsensical, then this initial category is deleted; If the differentiation feature of certain two initial category is roughly the same, and each differentiates that feature characteristic of correspondence value is also roughly the same, then will wherein any one initial category be merged in another initial category.Thus determine customer group classification final in customer group feature database.
By the customer group classification of acquisition, corresponding customer group, step S106, differentiates that feature and customer group differentiate that the reference value of feature is stored in customer group feature database.
As shown in Figure 4, in one embodiment, the detailed process of step S30 is:
Step S302, obtains the eigenwert in the characteristic information of user.
Such as, the characteristic information of certain user is: online class software 5 sections, online class software application frequency 1 hour/day, amusement class software 3 sections, amusement class software application frequency 2 hours/day, tool-class software 2 sections, tool-class software application frequency 0.5 hour/day, then can eigenwert 5,1,3,2,2,0.5 in characteristic information extraction.Further, the eigenwert composition proper vector can will extracted.As above in example, eigenwert is formed proper vector for (5,1,3,2,2,0.5).
Step S304, calculates the distance between reference value that the eigenwert customer group not corresponding with all types of user realm differentiate feature.
In one embodiment, the Euclidean distance between reference value that the eigenwert of user and all types of user group differentiate feature is calculated.Concrete, respectively the eigenwert obtained in step S302 and customer group can be differentiated the reference value composition characteristic vector of feature, and calculate the Euclidean distance between two proper vectors.
In addition, when the dimension of user characteristics value vector is greater than the dimension of the reference vectors of the differentiation feature of some customer group classifications, then to differentiate that the reference vectors of feature is for benchmark, reject the redundant character value beyond the differentiation feature reference value in user characteristics value vector, make user characteristics value vector consistent with differentiation feature reference value vector dimension, then calculate Euclidean distance between the two.
Step S306, obtains the customer group classification that minor increment is corresponding, the customer group classification belonging to user.
The eigenwert of user and customer group differentiate that the distance of the reference value of feature is less, then illustrate that this user customer group classification corresponding with this differentiation feature is more close, the minimum customer group of both acquisitions distance differentiates the customer group classification of customer group classification belonging to user that feature is corresponding.
In another embodiment, above-mentioned software recommendation method also comprises step: the software recommend list that Servers installed is corresponding with customer group classification.Concrete, corresponding software recommend list can be set according to the differentiation feature of each customer group classification, and store the mapping table of customer group classification and software recommend list in the server.Such as, for game class user, software recommend list can be set and comprise: QQ driving, hero kill, QQ tri-state etc.The software recommend list of setting like this more can meet the potential demand of user, improves the probability that user accepts to recommend software, thus improves the success ratio of software recommend further.
As shown in Figure 5, in one embodiment, the detailed process of above-mentioned steps S40 is:
Step S402, chooses corresponding software recommend list according to the customer group classification determined.
Customer group classification belonging to user, finds corresponding software recommend list in the mapping table of customer group classification to software recommend list.
Step S404, filter out from software recommend list user uninstalled or than the software recommend of the software version update of user installation to user.
Concrete, the current software matrix installed of user can be obtained according to the characteristic information of user, the software matrix installed and software recommend list are contrasted, the software of the uninstalled software of user and software version update more mounted than user in software recommend list can be obtained, then the software of the uninstalled software of user and software version update more mounted than user in the software recommend list obtained more is recommended user.In addition, after client receives the software of server recommendation, show, user can select corresponding software to carry out downloading and installing.
As shown in Figure 6, in one embodiment, a kind of software recommend system, comprises server 100 and carries out the client 200 of data interaction with server 100; Server 100 comprises creation module 110, matching module 120 and recommending module 130; Client 200 comprises information extraction modules 210; Wherein:
Creation module 110, for creating customer group feature database, stores customer group classification and corresponding customer group differentiation feature in customer group feature database.
Customer group classification carries out the rear plurality of classes obtained of classifying, as game user class, User class etc. to a large amount of users.Customer group differentiates that feature refers to the feature for differentiating the customer group classification belonging to user, such as, the differentiation feature of User class can be install the number of online class software, the number of the amusement class software of installation, the study class software of installation number, use time period etc. of software.
The characteristic information of user for extracting the characteristic information of user, and is submitted to server 100 by information extraction modules 210.
Use user in the process of various software, the raw information of client 200 recording user, such as, filled software matrix, the set-up time of software, the frequency of utilization of software, the upgrading frequency of software, user browse webpage record etc.Information extraction modules 210 characteristic information extraction from the raw information of record of client 200, the characteristic information extracted can be used for determining that user's owning user realm is other.Concrete, characteristic information comprises all kinds of software number, software set-up time, all kinds of software application frequency, all kinds of software upgrading frequency, all kinds of web page browsing durations etc., and wherein, the classification of software can be online class, amusement class, game class, study class etc.
In one embodiment, information extraction modules 210 can be used for extracting from the raw information of user for determining other characteristic information of user's owning user realm, and the characteristic information of extraction is submitted to server 100.In another embodiment, information extraction modules 210 also can be used for directly the raw information of user being submitted to server 100, by server 100 characteristic information extraction from the raw information of user.
Matching module 120, for the characteristic information of user and the customer group in customer group feature database are differentiated that feature is mated, determines the customer group classification belonging to user.
Concrete, matching module 120 differentiates feature for obtaining the customer group of mating most with user's characteristic information, then corresponding customer group classification is the customer group classification belonging to user.
Recommending module 130 is for recommending software according to the customer group classification determined to user.
Different customer group classifications can be arranged recommends different software, and such as, for game user class, then recommended games class software, for office users class, then recommends office class software.
Above-mentioned software recommend system, differentiate that feature is mated by the characteristic information of user client 200 extracted with the customer group in the customer group feature database in server 100, determine the customer group classification belonging to user, right rear line recommends the software corresponding with customer group classification.Software interested to different customer group classifications is different, according to the software that the characteristic information of user recommends its owning user realm not corresponding to user, can improve the possibility that user accepts to recommend software, thus improve the success ratio of software recommend.
As shown in Figure 7, in one embodiment, creation module 110 comprises sample and chooses module 112, preliminary classification module 114, characteristic information acquisition module 116 and feature database establishment submodule 118, wherein:
Sample chooses module 112 for choosing sample of users.
Concrete, sample chooses module 112 for selected part user from the user data stored as sample of users, the sample of users chosen relates to different classifications as far as possible, as network bar users class, User class, child user class, the elderly's user class, software development user class, game user class etc.
Preliminary classification module 114, for classifying to described sample of users, obtains the initial category of described sample of users.
Characteristic information acquisition module 116 is for obtaining the characteristic information of sample of users.
The user data stored in server 100 comprises the raw information of user, therefore need extract the characteristic information of user from raw information.
Preferably, characteristic information acquisition module 116, for obtaining the user profile of sample of users, extracts characteristic information relevant to software in user profile.Concrete, characteristic information acquisition module 116 can be used for by traditional Euclidean distance tolerance, by probability metrics criterion, with divergence criterion function, to combine extract the characteristic information relevant to software based on one or more in the methods such as differentiation entropy minimization.The characteristic information relevant to software such as extracted comprise the number of all kinds of softwares of installation, the set-up time of all kinds of software, the frequency of utilization of all kinds of software, the upgrading frequency of all kinds of software and all kinds of webpage browse duration etc.
Feature database creates submodule 118 and creates customer group feature database for the characteristic information according to initial category and user.
As shown in Figure 8, in one embodiment, feature database creates submodule 118 and comprises Cluster Analysis module 1180, customer group classification acquisition module 1182 and customer group feature database 1184, wherein:
Cluster Analysis module 1180 is for carrying out cluster analysis according to initial category to the characteristic information of user.
Concrete, Cluster Analysis module 1180 can be used for adopting traditional C mean algorithm, ISODATA algorithm, combines carry out cluster analysis to the characteristic information of user based on any one or a few method in the method such as similarity measurements quantity algorithm, adjacent function criterion algorithm of sample and core.
Customer group classification acquisition module 1182 differentiates the reference value of feature for obtaining customer group classification, corresponding customer group differentiation feature and described customer group according to cluster analysis result.
If for the some features in characteristic information, the corresponding eigenwert of all sample of users of a certain initial category all relatively, and differ greatly with the corresponding eigenwert of the sample of users of other initial category, illustrate that this feature not only has cohesion to this customer group, also this customer group and other customer group are separated by this feature, then this is characterized as the differentiation feature of this customer group.Contrary, if for the sample of users of some initial category, there is very big-difference in the eigenwert of its some feature, then as this initial user realm, other does not differentiate feature to this feature.Such as in 100 features of game user class, there is very large difference in the eigenwert of this feature of installation number of tool-class software, then this feature of installation number of tool-class software is not as the differentiation feature of game user class.
Further, customer group classification acquisition module 1182 can be used for the reference value determining this differentiation feature according to the corresponding eigenwert of all sample of users of this initial category.Concrete, customer group classification acquisition module 1182 can be used for getting the reference value of the mode in the mean value of all eigenwerts corresponding to this differentiation feature or all eigenwerts as this differentiation feature of this customer group.
Further, customer group classification acquisition module 1182 can be used for determining final customer group classification, inappropriate preliminary classification is deleted or merges.If for each feature in characteristic information, between the sample of users eigenwert of a certain initial category, all there is very large difference, illustrate that the division of this initial category is nonsensical, then this initial category is deleted; If the differentiation feature of certain two initial category is roughly the same, and each differentiates that feature characteristic of correspondence value is also roughly the same, then will wherein any one initial category be merged in another initial category.Thus determine the customer group classification that customer group feature database is final.
Customer group feature database 1184 differentiates the reference value of feature for storing customer group classification, corresponding customer group differentiation feature and described customer group.
As shown in Figure 9, in one embodiment, matching module 120 comprises characteristic value acquisition module 122, distance calculation module 124 and class of subscriber determination module 126, wherein:
Characteristic value acquisition module 122 is for obtaining the eigenwert in the characteristic information of described user.
Such as, the characteristic information of certain user is: online class software 5 sections, online class software application frequency 1 hour/day, amusement class software 3 sections, amusement class software application frequency 2 hours/day, tool-class software 2 sections, tool-class software application frequency 0.5 hour/day, then can eigenwert 5,1,3,2,2,0.5 in characteristic information extraction.Further, the eigenwert that characteristic value acquisition module 122 can be used for extracting forms proper vector.As above in example, eigenwert is formed proper vector for (5,1,3,2,2,0.5).
Distance calculation module 124 differentiates the distance between the reference value of feature for calculating the eigenwert customer group not corresponding with all types of user realm.
In one embodiment, distance calculation module 124 differentiates the Euclidean distance between the reference value of feature for the eigenwert and all types of user group calculating user.Concrete, distance calculation module 124 can be used for the reference value composition characteristic vector of eigenwert and customer group differentiation feature characteristic value acquisition module 122 obtained respectively, and calculates the Euclidean distance between two proper vectors.
In addition, when the dimension of user characteristics value vector is greater than the dimension of the reference vectors of the differentiation feature of some customer group classifications, then to differentiate that the reference vectors of feature is for benchmark, reject the redundant character value beyond the differentiation feature reference value in user characteristics value vector, make user characteristics value vector consistent with differentiation feature reference value vector dimension, then calculate Euclidean distance between the two.
Class of subscriber determination module 126 is for obtaining customer group classification corresponding to minimum described distance, the customer group classification belonging to described user.
The eigenwert of user and customer group differentiate that the distance of the reference value of feature is less, then illustrate that this user customer group classification corresponding with this differentiation feature is more close, class of subscriber acquisition module 126 differentiates for obtaining the minimum customer group of both distances the customer group classification of customer group classification belonging to user that feature is corresponding.
In another embodiment, server 100 also comprises and arranges module (not shown), for arranging the software recommend list corresponding with customer group classification.
Concrete, module is set and can be used for arranging corresponding software recommend list according to the differentiation feature of each customer group classification, and in server 100, store the mapping table of customer group classification and software recommend list.Such as, for game class user, software recommend list can be set and comprise: QQ driving, hero kill, QQ tri-state etc.The software recommend list of setting like this more can meet the potential demand of user, improves the probability that user accepts to recommend software, thus improves the success ratio of software recommend.
As shown in Figure 10, in one embodiment, recommending module 130 comprises list and chooses module 132, screening module 134, wherein:
List chooses module 132 for choosing corresponding software recommend list according to the customer group classification determined.
Concrete, list selects module 132 for the customer group classification belonging to user, in the mapping table of customer group classification to software recommend list, find corresponding software recommend list.
Screening module 134 for filter out from software recommend list user uninstalled or than the software recommend of the software version update of user installation to user.
Concrete, screening module 134 is for obtaining the current software matrix installed of user according to the characteristic information of user, the software matrix installed and software recommend list are contrasted, the software of the uninstalled software of user and software version update more mounted than user in software recommend list can be obtained, then the software of the uninstalled software of user and software version update more mounted than user in the software recommend list obtained more is recommended user.In addition, after client 200 receives the software of server 100 recommendation, can show, user can select corresponding software to carry out downloading and installing.
The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (10)
1. a software recommendation method, is characterized in that, comprises the following steps:
Server creates customer group feature database, stores customer group classification and corresponding customer group differentiation feature in described customer group feature database;
Client extracts the characteristic information of user, and the characteristic information of described user is submitted to described server;
With the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated by described server, determine the customer group classification belonging to described user;
Described server recommends software according to the described customer group classification determined to described user.
2. software recommendation method according to claim 1, is characterized in that, the step that described server creates customer group feature database comprises:
Choose sample of users;
Described sample of users is classified, obtains the initial category of described sample of users;
Obtain the characteristic information of described sample of users;
Characteristic information according to described initial category and user creates customer group feature database.
3. software recommendation method according to claim 2, is characterized in that, the step of the characteristic information of described acquisition sample of users is:
Obtain the user profile of sample of users, extract characteristic information relevant to software in described user profile;
The step that the described characteristic information according to described initial category and user creates customer group feature database is:
According to described initial category, cluster analysis is carried out to the characteristic information of described user;
Obtain customer group classification according to cluster analysis result, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature;
Described customer group classification, corresponding customer group are differentiated that feature and described customer group differentiate that the reference value of feature is stored in customer group feature database.
4. software recommendation method according to claim 3, is characterized in that, with the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated by described server, determines that the step of the customer group classification belonging to described user is:
Obtain the eigenwert in the characteristic information of described user;
Calculate the distance between reference value that the described eigenwert described customer group not corresponding with all types of user realm differentiate feature;
Obtain the customer group classification that minimum described distance is corresponding, the customer group classification belonging to described user.
5. software recommendation method according to claim 1, is characterized in that, described method also comprises:
The software recommend list that described Servers installed is corresponding with described customer group classification;
Described server recommends the step of software to be according to the described customer group classification determined to described user:
Corresponding software recommend list is chosen according to the described customer group classification determined;
Filter out from described software recommend list user uninstalled or than the software version update of user installation software recommend give described user.
6. a software recommend system, is characterized in that, comprise server and carry out the client of data interaction with described server, described server comprises:
Creation module, for creating customer group feature database, stores customer group classification and corresponding customer group differentiation feature in described customer group feature database;
Described client comprises: information extraction modules, for extracting the characteristic information of user, and the characteristic information of described user is submitted to described server;
Described server also comprises:
Matching module, for the customer group in described customer group feature database, the characteristic information of described user is differentiated that feature is mated, determines the customer group classification belonging to described user;
Recommending module, for recommending software according to the described customer group classification determined to described user.
7. software recommend system according to claim 6, is characterized in that, described creation module comprises:
Module chosen by sample, for choosing sample of users;
Preliminary classification module, for classifying to described sample of users, obtains the initial category of described sample of users;
Characteristic information acquisition module, for obtaining the characteristic information of described sample of users;
Feature database creates submodule, creates customer group feature database for the characteristic information according to described initial category and user.
8. software recommend system according to claim 7, is characterized in that, described characteristic information acquisition module also for obtaining the user profile of described sample of users, extracts characteristic information relevant to software in described user profile;
Described feature database creates submodule and comprises:
Cluster Analysis module, for carrying out cluster analysis according to described initial category to the characteristic information of described user;
Customer group classification acquisition module, for obtaining customer group classification according to cluster analysis result, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature;
Customer group feature database, for storing described customer group classification, corresponding customer group differentiates feature and the reference value of described customer group differentiation feature.
9. software recommend system according to claim 8, is characterized in that, described matching module comprises:
Characteristic value acquisition module, for obtaining the eigenwert in the characteristic information of described user;
Distance calculation module, the distance between the reference value differentiating feature for calculating described eigenwert and all kinds of described customer group;
Class of subscriber determination module, described apart from minimum customer group classification for obtaining, the customer group classification belonging to described user.
10. software recommend system according to claim 6, is characterized in that, described server also comprises:
Module is set, for arranging the software recommend list corresponding with described customer group classification;
Described recommending module comprises:
Module is chosen in list, for choosing corresponding software recommend list according to the described customer group classification determined;
Screening module, for filter out from described software recommend list described user uninstalled or than the software version update of user installation software recommend give described user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110304854.5A CN103034508B (en) | 2011-10-10 | 2011-10-10 | Software recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110304854.5A CN103034508B (en) | 2011-10-10 | 2011-10-10 | Software recommendation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103034508A CN103034508A (en) | 2013-04-10 |
| CN103034508B true CN103034508B (en) | 2015-08-19 |
Family
ID=48021434
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110304854.5A Active CN103034508B (en) | 2011-10-10 | 2011-10-10 | Software recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103034508B (en) |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103200279B (en) * | 2013-04-28 | 2017-03-15 | 百度在线网络技术（北京）有限公司 | Recommendation method and cloud server |
| CN104123315B (en) * | 2013-04-28 | 2018-01-30 | 北京音之邦文化科技有限公司 | The recommendation method and recommendation server of multimedia file |
| CN103327102B (en) * | 2013-06-24 | 2016-12-28 | 小米科技有限责任公司 | A kind of method and apparatus recommending application program |
| CN104615452A (en) * | 2014-09-19 | 2015-05-13 | 安一恒通（北京）科技有限公司 | Information providing method and device |
| CN105653833B (en) * | 2014-11-12 | 2019-04-26 | 腾讯科技（深圳）有限公司 | A kind of method and device that game community is recommended |
| CN105825391A (en) * | 2015-01-06 | 2016-08-03 | 中国移动通信集团公司 | Method and equipment for determining user attributes |
| CN104793932A (en) * | 2015-03-24 | 2015-07-22 | 百度在线网络技术（北京）有限公司 | Method and device used for software release |
| CN104869529A (en) * | 2015-04-22 | 2015-08-26 | 惠州Tcl移动通信有限公司 | Mobile terminal, server, and information management methods thereof |
| CN106557938A (en) * | 2015-09-24 | 2017-04-05 | 网易（杭州）网络有限公司 | The method for pushing of virtual objects in a kind of game |
| CN105468762B (en) * | 2015-12-02 | 2019-10-25 | 上海斐讯数据通信技术有限公司 | A kind of push APP application/desktop theme method and system |
| CN106056271B (en) * | 2016-05-17 | 2020-04-28 | 珠海许继芝电网自动化有限公司 | Intelligent control method for user group power load response |
| CN107807935B (en) * | 2016-09-09 | 2019-05-07 | 腾讯科技（深圳）有限公司 | Using recommended method and device |
| CN107885742B (en) * | 2016-09-29 | 2021-09-07 | 腾讯科技（深圳）有限公司 | Service recommendation method and device |
| CN107886345B (en) * | 2016-09-30 | 2021-12-07 | 阿里巴巴集团控股有限公司 | Method and device for selecting data object |
| CN108073626B (en) * | 2016-11-15 | 2020-06-19 | 平安科技（深圳）有限公司 | Target customer group positioning method and device |
| CN108074122A (en) * | 2016-11-18 | 2018-05-25 | 腾讯科技（深圳）有限公司 | Product beta test recommends method, apparatus and server |
| CN107704494B (en) * | 2017-08-24 | 2021-09-14 | 深圳市来玩科技有限公司 | User information collection method and system based on application software |
| CN108109052A (en) * | 2017-12-29 | 2018-06-01 | 广州优视网络科技有限公司 | Article method for pushing, device and server |
| CN108510402A (en) * | 2018-06-06 | 2018-09-07 | 中国平安人寿保险股份有限公司 | Insurance kind information recommendation method, device, computer equipment and storage medium |
| CN111859117B (en) * | 2020-06-19 | 2024-03-26 | 腾讯科技（深圳）有限公司 | Information recommendation method and device, electronic equipment and readable storage medium |
| CN111736876A (en) * | 2020-06-28 | 2020-10-02 | 北京云族佳科技有限公司 | Client update control method, device, storage medium and server |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004206671A (en) * | 2002-11-08 | 2004-07-22 | Qwanty Corp | Software update information provision method and software update information provision system |
| CN101867594A (en) * | 2010-03-05 | 2010-10-20 | 宇龙计算机通信科技(深圳)有限公司 | Data transmission method, device and system |
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
-
2011
- 2011-10-10 CN CN201110304854.5A patent/CN103034508B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004206671A (en) * | 2002-11-08 | 2004-07-22 | Qwanty Corp | Software update information provision method and software update information provision system |
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN101867594A (en) * | 2010-03-05 | 2010-10-20 | 宇龙计算机通信科技(深圳)有限公司 | Data transmission method, device and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103034508A (en) | 2013-04-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103034508B (en) | Software recommendation method and system | |
| CN104090888B (en) | A kind of analytical method of user behavior data and device | |
| US8555173B2 (en) | Recommendation engine | |
| US20190147246A1 (en) | System and method for providing augmented reality interactions over printed media | |
| CN105975472A (en) | Method and device for recommendation | |
| CN106055617A (en) | Data pushing method and device | |
| US20120323725A1 (en) | Systems and methods for supplementing content-based attributes with collaborative rating attributes for recommending or filtering items | |
| US9088811B2 (en) | Information providing system, information providing method, information providing device, program, and information storage medium | |
| CN107818166A (en) | A kind of information recommends method, apparatus, server and system | |
| CN101681372A (en) | Method and system for providing relevant information to a user of a device in a local network | |
| CN103686382A (en) | Program recommendation method | |
| CN102523511A (en) | Network program aggregation and recommendation system and network program aggregation and recommendation method | |
| CN104063801A (en) | Mobile advertisement recommendation method based on cluster | |
| CN108650558A (en) | The generation method and device of video Previously on Desperate Housewives based on interactive video | |
| CN101103349A (en) | Method for extracting content, content extraction server based on RSS and apparatus for managing the same and system for providing standby screen of mobile communication terminal using the same | |
| CN103106208A (en) | Streaming media content recommendation method and system in mobile internet | |
| CN105653572A (en) | Resource processing method and apparatus | |
| CN104662529A (en) | Data refining engine for high performance analysis system and method | |
| CN103198086A (en) | Information processing device, information processing method, and program | |
| CN103761228B (en) | The rank threshold of application program determines that method and rank threshold determine system | |
| CN106202393B (en) | Media information pushing method and device | |
| US20130212105A1 (en) | Information processing apparatus, information processing method, and program | |
| CN102129444A (en) | Information processing apparatus, information processing method, and program | |
| CN109829063A (en) | A kind of data processing method, device and storage medium | |
| KR101870980B1 (en) | System for recommend the customized application, method thereof and recordable medium storing the method |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190808 Address after: 518000 Nanshan District science and technology zone, Guangdong, Zhejiang Province, science and technology in the Tencent Building on the 1st floor of the 35 layer Co-patentee after: Tencent cloud computing (Beijing) limited liability company Patentee after: Tencent Technology (Shenzhen) Co., Ltd. Address before: Shenzhen Futian District City, Guangdong province 518044 Zhenxing Road, SEG Science Park 2 East Room 403 Patentee before: Tencent Technology (Shenzhen) Co., Ltd. |