CN102567511A

CN102567511A - Method and device for automatically recommending application

Info

Publication number: CN102567511A
Application number: CN2011104447985A
Authority: CN
Inventors: 叶松; 秦吉胜; 常富洋
Original assignee: Qizhi Software Beijing Co Ltd
Current assignee: Qizhi Software Beijing Co Ltd
Priority date: 2011-12-27
Filing date: 2011-12-27
Publication date: 2012-07-11
Anticipated expiration: 2031-12-27
Also published as: CN102567511B

Abstract

The invention provides a method and a device for automatically recommending an application. The method comprises the steps of: receiving an application acquisition request submitted from a client by a user, wherein the application acquisition request comprises a user identification; according to the user identification, extracting the existing user behavior information of corresponding user from a user feature library, wherein the user behavior information comprises operation information that a user aims at previously recommended application; according to the user behavior information, determining the type of the application recommended to the user; in an application data set of the type of the application, according to the operation information that the user aims at the previously recommended application, extracting an matched application; and according to the type of the application, generating a corresponding application folder, and placing the matched application in the corresponding application folder for recommending. According to the invention, the individual needs of users can be satisfied, the recommending efficiency is improved and the coverage rate is increased.

Description

A kind of method and device of using automatic recommendation

Technical field

The application relates to technical field of information processing, particularly relates to a kind of method and automatic recommended device of a kind of application of using automatic recommendation.

Background technology

The internet is the important channel that people obtain information; When the principal feature of tradition internet is that the user seeks own interested things, need carry out a large amount of search, need the artificially to filter out a large amount of incoherent results simultaneously through browser; Complex operation, and expend time in and energy.

Develop rapidly along with Internet technology; The demand that people use (Application) to diverse network also more and more widely; But along with the increase of demand, the terminal applies that people install in terminal clientsaconnect is also more and more, and the various deployment that are applied in client are more and more too fat to move huge; This not only causes the waste to terminal resource, nor is convenient to management.Even adopt client-server architecture to dispose management, server end also lacks the managerial ability to follow-up use after the deployment of accomplishing client.

Although occurred the notion of so-called " thin-client (Thin Client) " now, thin-client is sent to server process with inputs such as its mouse, keyboards, and server is back to client to result again and shows.But this tupe is limited by network transfer speeds, and the restrictions such as processing power of server, therefore, more is to be applied in the commercial LAN of enterprise-level, also is not suitable for the amusement demand of domestic consumer at present.

For making the user obtain better experience, prior art has proposed the scheme that provides interested application to recommend automatically for the user, promptly through knowing user's interest place, is its recommendation initiatively, its interested application is provided.Yet, the mode that this application is recommended, main all through the manual recommendation of editorial staff, mainly there is following defective in the manual mode of recommending of this editorial staff:

1, efficient is low excessively, and is too low for the recommendation coverage rate of using, and for example, for application hundreds thousand of on the platform, adopts artificial the recommendation every day, also can only recommend hundreds of individual.If want to recommend whole application in fact can't realize, and coverage rate is too low, because proportion is too low.

2, this recommendation unified principle of recommending that is based on the editorial staff is fully carried out, and is all the same for each user, can't satisfy the demand of user individual.Because the application of some recommendation is suitable for the certain user, and dislikes for the certain user.

Therefore, need the urgent technical matters that solves of those skilled in the art to be exactly at present: to propose a kind of mechanism of using automatic recommendation, satisfying user's individual demand, and improve recommendation efficient and coverage rate.

Summary of the invention

The application's technical matters to be solved provides a kind of method of using automatic recommendation, in order to satisfying user's individual demand, and improves and recommends efficient and coverage rate.

The application also provides a kind of application automatic recommended device, in order to guarantee application and the realization of said method in reality.

In order to address the above problem, the application embodiment discloses a kind of method of using automatic recommendation, comprising:

Receive the user and obtain request from the application of client submission, said application is obtained and is comprised ID in the request;

From the user characteristics storehouse, extract the existing user behavior information of relative users according to said ID, said user behavior information comprises that the user is directed against the operation information of exemplary application formerly;

Confirm applicating category according to said user behavior information to user's recommendation;

In the application data sets of said applicating category, being directed against formerly according to the user, the operation information of exemplary application extracts matched application;

Generate corresponding application file folder by said applicating category, said matched application is put into corresponding application file folder recommend.

Preferably, described method can also comprise:

Gather the user behavior information of submitting to after request is obtained in said application, write in the user characteristics storehouse by ID.

Preferably, said user behavior information also comprises user's local operation behavioural information, and/or, user's online operation behavior information;

The step of the said applicating category of confirming according to user behavior information to recommend to the user can comprise:

From said user's local operation behavioural information and/or online operation behavior information, extract the tag along sort and the first corresponding operation frequency;

Convert said tag along sort into corresponding applicating category by preset correlation rule; Said preset correlation rule is the transformation rule of tag along sort and applicating category;

From the operation information of said user to exemplary application formerly, extract user operated application message and second corresponding operation frequency in the Preset Time section, comprise applicating category in the said application message;

Calculate the weight of each applicating category according to the said first operation frequency and the second operation frequency, sort from high to low by the weight of said applicating category;

Preceding n the applicating category that extracts predetermined number is the applicating category of recommending to the user; Wherein, said n is the positive integer greater than 1.

Preferably, can generate the application data set of certain applicating category through following steps:

Obtain the application of same applicating category, said application has tag along sort;

In said application, confirm main the application and application to be recommended, and calculate application to be recommended and the main similarity of using according to the tag along sort of each application;

Obtain said application quality grading parameters to be recommended;

Extract the same main pairing application to be recommended of using respectively, sort from high to low, and extract the individual application to be recommended of m before the predetermined number by similarity and the quality score parameter of each application to be recommended; Wherein, said m is the positive integer greater than 1;

The application data set of the main application corresponding to be recommended of using and being extracted being formed the current application classification.

Preferably, said application data sets at applicating category, being directed against formerly according to the user, the step of the operation information extraction matched application of exemplary application can comprise:

Be directed against the operation information of exemplary application formerly according to the user, the 3rd of main application of statistics and correspondence operated the frequency, and said master is applied as the operated application of user;

Application data sets at corresponding applicating category; Application to be recommended according to said main application fetches coupling; And in the application to be recommended of said coupling; With the application to be recommended that said the 3rd operation frequency is extracted some respectively as the weight of application fetches, extract the application to be recommended of satisfying first predetermined number altogether.

Preferably, said application data sets at applicating category, being directed against formerly according to the user, the step of the operation information extraction matched application of exemplary application can also comprise:

Obtain the main corresponding applicating category of using, in same applicating category, said main the application sorted, extract preceding k main application of predetermined number by said the 3rd operation frequency; Wherein, said k is the positive integer greater than 1;

The main application in twos of being extracted matched, calculate the main total degree that occurs simultaneously of using of said pairing in twos, generate frequent 2 collection;

Calculate each main number of times that occurs separately of using, generate frequent 1 collection;

Calculate each main degree of confidence of using according to said frequent 2 collection and frequent 1 collection, and main the application sorted by degree of confidence;

With the application to be recommended of satisfying first predetermined number of being extracted, and said main application the by the degree of confidence ordering mated, and generates final coupling of recommending and uses.

The application embodiment discloses the automatic recommended device of a kind of application simultaneously, comprising:

The request receiver module is used to receive the user and obtains request from the application of client submission, and said application is obtained and comprised ID in the request;

Formerly the behavioural information extraction module is used for extracting the existing user behavior information of relative users from the user characteristics storehouse according to said ID, and said user behavior information comprises that the user is directed against the operation information of exemplary application formerly;

The applicating category determination module is used for confirming the applicating category to user's recommendation according to said user behavior information;

Coupling is used acquisition module, is used for the application data sets at said applicating category, and being directed against formerly according to the user, the operation information of exemplary application extracts matched application;

Use recommending module, be used for generating corresponding application file folder, said matched application is put into corresponding application file folder recommend by said applicating category.

Preferably, described device also comprises:

The behavioral statistics module is used to gather the user behavior information of submitting to after request is obtained in said application, writes in the user characteristics storehouse by ID.

Said applicating category determination module can comprise:

The first feature extraction submodule is used for local operation behavioural information and/or online operation behavior information from said user, extracts the tag along sort and the first corresponding operation frequency;

The conversion submodule is used for converting said tag along sort into corresponding applicating category by preset correlation rule; Said preset correlation rule is the transformation rule of tag along sort and applicating category;

The second feature extraction submodule is used for from said user to the operation information of exemplary application formerly, extracts user operated application message and second corresponding operation frequency in the Preset Time section, comprises applicating category in the said application message;

The ordering submodule is used for calculating according to the said first operation frequency and the second operation frequency weight of each applicating category, sorts from high to low by the weight of said applicating category;

Classification is selected submodule, is used to extract the applicating category of preceding n applicating category for recommending to the user of predetermined number; Wherein, said n is the positive integer greater than 1.

Preferably, described device can also comprise:

The application data set generation module is used to generate the application data set of each applicating category: specifically comprise:

Submodule is obtained in similar application, is used to obtain the application of same applicating category, and said application has tag along sort;

The similarity calculating sub module is used for confirming main the application and application to be recommended in said application, and calculates application to be recommended and the main similarity of using according to the tag along sort of each application;

Quality score parameter acquiring submodule is used to obtain said application quality grading parameters to be recommended;

Application fetches submodule to be recommended is used for extracting respectively the same main pairing application to be recommended of using, and sorts from high to low by similarity and the quality score parameter of each application to be recommended, and extracts the individual application to be recommended of m before the predetermined number; Wherein, said m is the positive integer greater than 1;

Application data set forms submodule, is used for the main application corresponding to be recommended of using and being extracted is formed the application data set of current application classification.

Preferably, said coupling application acquisition module can comprise:

Main applied statistics submodule is used for being directed against the operation information of exemplary application formerly according to the user, and the 3rd of main application of statistics and correspondence operated the frequency, and said master is applied as the operated application of user;

Submodule is confirmed in application to be recommended; Be used for application data sets at corresponding applicating category; Application to be recommended according to said main application fetches coupling; And in the application to be recommended of said coupling,, extract the application to be recommended of satisfying first predetermined number altogether with the application to be recommended that said the 3rd operation frequency is extracted some respectively as the weight of application fetches.

Preferably, said coupling application acquisition module can also comprise:

The main application chosen submodule, is used to obtain the main corresponding applicating category of using, and in same applicating category, by said the 3rd operation frequency said main the application sorted, and extracts preceding k main application of predetermined number; Wherein, said k is the positive integer greater than 1;

Frequent 2 collection calculating sub module are used for main the application in twos of being extracted matched, and calculate the main total degree that occurs simultaneously of using of said pairing in twos, generate frequent 2 collection;

Frequent 1 collection calculating sub module is used to calculate each main number of times that occurs separately of using, and generates frequent 1 collection;

The confidence calculations submodule is used for calculating each main degree of confidence of using according to said frequent 2 collection and frequent 1 collection, and by degree of confidence main the application is sorted;

Coupling is used and is confirmed submodule, be used for the application to be recommended of satisfying first predetermined number of being extracted, and said main application the by the degree of confidence ordering is mated, and generates final coupling of recommending and uses.

Compared with prior art, the application has the following advantages:

The application is based on the application of having recommended to the user; Analysis user in conjunction with user's online operation behavior information and/or local operation behavioural information, is confirmed the applicating category of user behavior information institute preference to the operation information of said exemplary application formerly; Then in the application data sets of corresponding applicating category; According to the operation information of above-mentioned user,, extract the application that meets user interest most in conjunction with user's online operation behavior information and/or local operation behavioural information to said exemplary application formerly; These application being put into the file of corresponding applicating category recommends; Thereby between application and user, set up contact, fully satisfied user's individual demand, and effectively improved recommendation efficient and the coverage rate used.

Moreover, the application with user interface as inlet, directly on the interface or through the link on the interface through the application file clip icon to user's exemplary application obtain required application so that the user is faster easier, made things convenient for user's operation; And, can point out the user use through icon as the mode of application entrance, but before the user really selected to use, actual installation should be used corresponding configuration file to this application, like this, can before use and exceed and take client resource.In addition, the icon in the user interface can be concentrated by the network side central server and dispose or push, and this has just prevented that rogue program from arbitrarily adding the malice icon in the interface, further improved security.

Description of drawings

Fig. 1 is a kind of flow chart of steps of using the method embodiment of automatic recommendation of the application;

Fig. 2 is a kind of structured flowchart of using automatic recommended device embodiment of the application.

Embodiment

For above-mentioned purpose, the feature and advantage that make the application can be more obviously understandable, the application is done further detailed explanation below in conjunction with accompanying drawing and embodiment.

The core idea of the application embodiment is that based on the application of having recommended to the user, analysis user is to the operation information of said exemplary application formerly; Online operation behavior information and/or local operation behavioural information in conjunction with the user; Confirm the applicating category of user behavior information institute preference,, be directed against the operation information of said exemplary application formerly according to above-mentioned user then in the application data sets of corresponding applicating category; Online operation behavior information and/or local operation behavioural information in conjunction with the user; Extract the application that meets user interest most, these application are put into the file of corresponding applicating category and recommend, thereby between application and user, set up contact.

With reference to Fig. 1, a kind of flow chart of steps of using the method embodiment of automatic recommendation that it shows the application specifically can comprise the steps:

Step 101, reception user obtain request from the application that client is submitted to, and said application is obtained and comprised ID in the request;

In concrete the realization, the user starts client can trigger the request of obtaining of using, and the user also can the manual triggers application obtain request, and the application does not limit this.

Step 102, from the user characteristics storehouse, extract the existing user behavior information of relative users according to said ID, said user behavior information comprises that the user is directed against the operation information of exemplary application formerly;

Said user characteristics can write down following information: ID Mid in the storehouse, the tag along sort tag of user behavior information, and, corresponding operation frequency weight.

In a kind of preferred embodiment of the application, said user's behavioural information can comprise user's local operation behavioural information, and/or, user's online operation behavior information, and the user is directed against the operation information of exemplary application formerly.Said user's local operation behavioural information can have tag along sort (tag) information usually with online operation behavior information; For example; For the video that the user opens at local operation, have classification label informations such as the fiery shadow person of bearing, animation, serial, illusion, risk, bank Ben Qishi; Or as, for the network address that the user is visited, have the classification label informations such as king of video, film, comedy movie, comedy on the net.Said application also has the information of applicating category and tag along sort.

Said user's local operation behavioural information can be gathered by the client software that is installed on the subscriber equipment with online operation behavior information; Wherein, said subscriber equipment can comprise all kinds of intelligent terminals such as computing machine, notebook computer, mobile phone, PDA, panel computer.Several kinds of collection users' local operation behavioural information below is provided, and/or, the example of user's online operation behavior information:

Example 1 is gathered the online operation behavior information of user in a period of time through browser, comprises network address and the corresponding access times of visit etc.;

Online operation behavior information as gathering through browser in the user 15 days is:

Example 2; Through being installed in the fail-safe software collection user's on the subscriber equipment local operation behavioural information; Online operation behavior information and local behavioural information as gathering through 360 net shields in the users 15 days are: open MPC and number of times thereof, open certain recreation and number of times thereof etc.

Certainly, the method for above-mentioned collection and the information of collection are all only as example, and it all is feasible that those skilled in the art adopt any mode to gather required user behavior information according to actual conditions, and the application embodiment need not this to limit.

Step 103, confirm the applicating category recommended to the user according to said user behavior information;

In a kind of preferred embodiment of the application, said step 103 specifically can comprise following substep:

Substep S11, from said user's local operation behavioural information and/or online operation behavior information, extract the tag along sort and the first corresponding operation frequency;

Substep S12, convert said tag along sort into corresponding applicating category by preset correlation rule; Said preset correlation rule is the transformation rule of tag along sort and applicating category;

Substep S13, from the operation information of said user to exemplary application formerly, extract user operated application message and second corresponding operation frequency in the Preset Time section, comprise applicating category in the said application message;

Substep S14, calculate the weight of each applicating category, sort from high to low by the weight of said applicating category according to the said first operation frequency and the second operation frequency;

Preceding n applicating category of substep S15, extraction predetermined number is the applicating category of recommending to the user; Wherein, said n is the positive integer greater than 1.

In reality, can through the analysis user behavioural information, obtain the applicating category of user behavior information conforms according to applicating category being set in advance by the technician.For example; The application file folder basic classification that is provided with in advance has 20; And through the analysis user behavioural information; It is unwanted for the active user that discovery has some basic classification, and the applicating category that then can divide user behavior information and belonged to is 3 or 5 of behavioural habits before being close to the users more.For example, video, recreation, education etc.

For making those skilled in the art understand the application better, below confirm process according to user behavior information to the applicating category of user's recommendation through a concrete example description:

Data Source:

(1) nearest 15 days net shield data: Data11;

(2) the application interpolation or the click logs of nearest 15 days user's desktop safe in utilization: Data12;

The data layout of Data11 is:

The ID Mid tag along sort Interest first operation frequency weight1

The data instance of Data11 is following:

0000175873530b93d848614a0c188c5b novel-dm?1；

000020218613d5fc8e05c314dba32956 comic-dm 4；

00002e3bb9037870973b328078971c98 4399-dm 1；

The data layout of Data12 is:

Original log

The data instance of Data12 is following:

123.97.168.210--[15/Oct/2011:23:00:01+0800]″GET/stat.html?type＝open&action＝yingyongdianji&fangshi＝2&Appid＝102000032&fenleiid＝4&from＝0&leixing＝2&style＝fullscreen&uid＝1&pid＝softmgr&m＝45c06dc58f5ccb64c162b646fcecc541&modulever＝1.4.0.1103&appver＝1.4.0.1103HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?7.0；Windows?NT?6.1；Trident/4.0；GTB7.1；SLCC2；.NET?CLR?2.0.50727；.NET?CLR?3.5.30729；.NET?CLR?3.0.30729；Media?Center?PC?6.0)″

Step1: will from net shield data Data11, extract tag along sort Interest; Convert the basic classification that application file presss from both sides the user interest under the taxonomic hierarchies into through preset conversion rule table (yunCatToZhuoMianCat.conf), be about to said type of label and convert corresponding applicating category into.Can comprise in the said preset conversion rule table yunCatToZhuoMianCat.conf form: the information of tag along sort, applicating category title AppName and applicating category sign Appid, as shown in the table:

The result of Data11 conversion is as shown in the table:

Mid	Interest	AppName	Appid	Weight1
					000020218613d5fc8e05c314dba32956	comic-dm	The fashion amusement	8	4
0000175873530b93d848614a0c188c5b	novel-dm	Novel	11	1
					00002e3bb9037870973b328078971c98	4399-dm	Recreation	5	1

Step2:, calculate each Mid clicked or added each application in nearest 15 days the operation frequency (the second operation frequency), and, confirm the pairing applicating category of user interest according to the Appid_name classification table of comparisons through resolving original log Data12.

Wherein, Appid_name classification table of comparisons form is following:

Application identities Appid applicating category sign fenleiid tag along sort tag Apply Names AppName.

The data instance of the Appid_name classification table of comparisons is following:

The single tractor of other chess and card cartoon English of 100,026,002 5 trivial games

100,013,330 4 TV play ancient costume story of a play or opera palace schemings

100,114,314 6 healthy home doctor autodiagnosis family doctors

The cook way of pork braised in brown sauce of 100,114,370 6 cuisines pork braised in brown sauce menu recipes

The military suspense confidential informant of 100,013,349 4 TV plays

If through resolving original log Data12, the data of calculating each Mid clicked or added each application in nearest 15 days the operation frequency (second operates frequency Weight2) are as shown in the table:

Contrast the above-mentioned Appid_name classification table of comparisons, confirm that the pairing applicating category of user interest is as shown in the table:

Mid	fenleiid	Weight2
			00008fc5c27c3354e1e0c9b6b7527dd9	5	1
0000b5d11c0c8ea46817fc32f467c3ba	4	3
			0001555e4ea2b299b6fbc55f46eeb771	6	4

Step3: carry out weighted mean to the result of Step1 and Step2 according to the first operation frequency and the second operation frequency; Sort according to final score then; Get the applicating category of top9, i.e. the applicating category of the final classification application file of showing for recommending to the user.

For example: for some Mid, the result of Step1 is: type1 clicks n1 time, and type2 clicks n2 time, and type3 clicks n3 time ...;

Step2 result is: type1 behavior N1 time, type2 clicks N2 time, type3 click N3 time ... score1=n1*0.6+N1*0.4 then, score2=n2*0.6+N2*0.4, score3=n3*0.6+N3*0.4......

Sort by score, get the applicating category of preceding 9 applicating category for recommending to the active user.

In concrete the realization; If being analyzed the applicating category of being divided, user behavior information can't reach specified quantity; If example can only generate three classifications in the employing; Can't satisfy the demand of 9 applicating categories, then can carry out polishing as the applicating category of recommending according to the applicating category of maximum applicating category of the actual access times of the network user that high in the clouds is added up or up-to-date setting.

Certainly; The method of above-mentioned division user behavior information institute belonging kinds is only as example; It all is feasible that those skilled in the art adopt a kind of mode according to actual conditions; For example, do not extract the Main classification label, directly with user behavior information with label convert applicating category into according to presetting rule; Perhaps, directly extract tag along sort as applicating category etc., the application does not limit this.

Step 104, in the application data sets of said applicating category, extract matched application according to the user to the operation information of exemplary application formerly;

Said application (Application) is meant user's employed various services on network, like application program, webpage, video, novel, music, recreation, news, shopping and mailbox etc.Application data set comprises a plurality of application, derives from each open platform.In the application embodiment, application can be with classification information (applicating category) and some tag along sorts.

In a kind of preferred embodiment of application, can generate the application data set of certain applicating category through following substep:

Substep S21, the application of obtaining same applicating category, said application has tag along sort;

Substep S22, in said application, confirm main the application and application to be recommended, and calculate application to be recommended and the main similarity of using according to the tag along sort of each application;

Substep S23, obtain said application quality grading parameters to be recommended;

Substep S24, extract the same main pairing application to be recommended of using respectively, sort from high to low, and extract the individual application to be recommended of m before the predetermined number by similarity and the quality score parameter of each application to be recommended; Wherein, said m is the positive integer greater than 1;

Substep S25, will lead the application data set that the current application classification is formed in the application corresponding to be recommended of using and being extracted.

Above-mentioned preferred embodiment promptly to the application of same applicating category, according to the similarity between its tag along sort computing application, forms an application data set that comprises main application and application to be recommended.

For making those skilled in the art understand the application better, below pass through the process of an above-mentioned generation application data set of concrete example description.

1). according to the similarity between the tag along sort tag computing application app of applicating category of using and application:

The data layout of input data is: Appid fenleiID tag1 tag2 tag3 tag4 tag5 tag6 tag7 tag8......;

The data layout of destination file Data1 is: main application identities (main Appid) application identities to be recommended (Appid to be recommended) similarity Similarity

Similarity calculating method is:

In identical applicating category, a former i tag is as the class mark, and in identical mark, app makes up in twos, calculates its similarity, and computing formula is: Similarity=i/ (n1+n2-i); Wherein, n1 is the number of app1 (application 1) back tag, and n2 is the number of app2 (application 2) back tag; The i minimum is 2, is n1 to the maximum, carries out searching loop.For example:

The input data are:

100,030,071 4 other 2010 continents of film story of a play or opera comedy love monarch Wu Chen Liu Yanjun Xie Xiaoming

Shining other 2009 Korea S of Shen Tailuo of 100,030,073 4 film story of a play or opera comedies action Jin Henajiang will

Other 2009 U.S. of the triumphant strange money De Lekanteburui Alex Pu Luoyasi of the terrible Nicholas of 100,030,074 4 film suspense science fictions

......

Destination file Data1 is:

100030071 100030073 0.25

100030071 100030074 0.11

100030073 100030074 0.11

......

2). in same main Appid; Undertaken integrated ordered to Appid to be recommended by the quality score of similarity and Appid (every day download, user's scoring); Be Appid similarity * similarity weight+Appid quality score * (1-similarity weight); Preceding 50 Appid to be recommended that the intercepting integrate score is the highest merge into delegation then;

Input data: Data1 (destination file in a last step)

The form of output data Data2 is following: main Appid Appid1 to be recommended Appid2 to be recommended Appid3 to be recommended Appid4...... to be recommended

For example:

Input data Data1 is:

100030071 100030073 0.25

100030071 100030074 0.11

100030073 100030074 0.11

100030073 100030071 0.25

100030074 100030071 0.11

100030074 100030073 0.11

......

Output data Data2 is:

100030071 100030073 100030074......

100030073 100030071 100030074......

100030074 100030071 100030073......

In a kind of preferred embodiment of application, said step 104 may further include following substep:

Substep S31, according to the user to the operation information of exemplary application formerly, statistics is main to be used and the 3rd corresponding operation frequency, said master is applied as the operated application of user;

Substep S32, in the application data sets of corresponding applicating category; Application to be recommended according to said main application fetches coupling; And in the application to be recommended of said coupling; With the application to be recommended that said the 3rd operation frequency is extracted some respectively as the weight of application fetches, extract the application to be recommended of satisfying first predetermined number altogether.

For making those skilled in the art understand the application better, below through the above-mentioned substep S31-S32 of concrete example description.

Step1：

3). according to the application operating behavior daily record of Mid, statistics Mid adds or clicks the number of times of each app;

Input data: Mid adds or clicks the log record (nearest 30 days application operating behavior daily record) of application;

The form of output data Data3 is: Mid master Appid (clicking the id of the app that perhaps adds)

Weight3 (the 3rd operation frequency)

If the input data are:

27.185.166.230--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝yingyongdianji&fangshi＝2&Appid＝103352&fenleiid＝10001&from＝0&leixing＝1&style＝fullscreen&uid＝1&pid＝h_home_inst&m＝71ddd8f9f1c84e16438ef109f4b6d77b&modulever＝1.4.0.1041&appver＝1.4.0.1041?HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?7.0；Windows?NT?6.0；SLCC1；.NET?CLR2.0.50727；Media?Center?PC?5.0；.NET?CLR?3.5.30729；.NET?CLR?3.0.30618)″

111.127.218.150--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝tianjiayingyong&Appid＝100018815&fenleiid＝4&sort＝％b6％af％bb％ad&from＝5&style＝fullscreen&uid＝1&pid＝h_home_inst&m＝9e236bafe13c8348247781c2d0fab7a7&modulever＝1.0.2.1025&appver＝1.4.0.1040?HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?6.0；Windows?NT?5.1；SV1；4399Box.909)″

58.50.201.130--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝yingyongdianji&fangshi＝2&Appid＝103352&fenleiid＝10001&from＝6&leixing＝1&style＝iphone&uid＝1&pid＝h_home&m＝3d3e77348ff2fbfa6af7c3751a00edae&modulever＝1.4.0.1040&appver＝1.4.0.1040HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?6.0；Windows?NT?5.1)″

110.178.40.7--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝yingyongdianji&fangshi＝2&Appid＝100000525&fenleiid＝4&from＝0&leixing＝2&style＝fullscreen&uid＝1&pid＝softmgr&m＝f16b5a2c01d64fcfa3ff5f035ce74677&modulever＝1.4.0.1041&appver＝1.4.0.1041?HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?7.0；Windows?NT?5.1；Trident/4.0)″

58.50.201.130--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝zuixiaohuazhuomian&count＝90468&uid＝1&pid＝h_home&m＝3d3e77348ff2fbfa6af7c3751a00edae&modulever＝1.4.0.1040&appver＝1.4.0.1040HTTP/1.1″200?0″-″″Mozilla/4.0(compatible；MSIE?6.0；Windows?NT?5.1)″

......

Output data Data3 is:

0004a218f3b8a96b59f67a8f14be5e98 100000289 102

0004ce91dc7726afdc420a97ad050f5a 100103087 1

0004cfe7c72aa83bfa20e561b6a00823 102020214 2

0004ec5290131cf7fa6d5a18833e06a5 102005903 3

0004f55c24b9fb8745b6d7595d94972d 120033743 1

00051004579f67de2066dc94f8952fd4 100000275 5

00054c7ba3cf1b45330444cbb737cac4 100114758 4

......

4). according to Data2 and Data3, mate, sort according to weight then through main app

Mid1?fenleiid1?Appid1?Appid2?Appid3?Appid4......weight

Mid1?fenleiid2?Appid11?Appid22?Appid33?Appid44......weight

According to the weight size, weight is big more, and app is many more in such the inside intercepting, adopts the mode of intercepting at random, and 50 Appid of intercepting are that key word merges with Mid and fenleiid then altogether, generate destination file to do

Mid1?fenleiid?Appid1?Appid2?Appid3?Appid4......Appid20

For example:

Input data: Data2 and Data3

Wherein, Data2 is:

001(100?101?102?103?104?105?106?107?108?109......)

002(201?202?203?204?205?206?207?208?209?210......)

003(301?302?303?304?305?306?307?308?309?310......)

008(801?802?803?804?805?806?807?808?809?810......)

......

Suppose that the part in the above-mentioned Data2 bracket is preceding 50 Appids the highest with main app similarity

Data3 is:

Xx1?00?15

Xx1?002?3

Xx1?003?2

Xx1?008?5

.......

Through the Appid_name classification table of comparisons, shine upon its classification, acquisition Data33 is:

Xx1?1 001?5

Xx1?1?002 3

Xx1?1?003?2

Xx1?2?008?5

......

Intermediate output data:

Article 1, data: Xx1 1 100 101 102 103 104 105 106 107 108 109......5

Article 2, data: Xx1 1 201 202 203 204 205 206 207 208 209 210......3

Article 3, data: Xx1 1 301 302 303 304 305 306 307 308 309 310......2

Article 4, data: Xx1 2 801 802 803 804 805 806 807 808 809 810......5

......

According to weighing computation method:

Article 1, randomly draw 20=10 Appid of 5/ (5+3+2) * in the data;

Article 2, randomly draw 20=6 Appid of 3/ (5+3+2) * in the data;

Article 3, randomly draw 20=4 Appid of 2/ (5+3+2) * in the data;

Final output data is:

Xx1?1?100?101?102?103?104?105?106?107?108?109?204?205?206?207?208?209306?307?308?303

Xx2?2?801?802?803?804?805?806?807?808?809?810......

......

More preferably, said step 104 may further include following substep:

Substep S33, obtain the main corresponding applicating category of using, in same applicating category, said main the application sorted, extract preceding k main application of predetermined number by said the 3rd operation frequency; Wherein, said k is the positive integer greater than 1;

Substep S34, main used pairing in twos, calculates the main total degree of appearance simultaneously of using of said pairing in twos, generate frequent 2 collection what extract;

Substep S35, calculate each main number of times that occurs separately of using, generate frequent 1 collection;

Substep S36, calculate each main degree of confidence of using, and main the application sorted by degree of confidence according to said frequent 2 collection and frequent 1 collection;

Substep S37, with the application to be recommended of satisfying first predetermined number of being extracted, and said main application the by the degree of confidence ordering mated, and generates final coupling of recommending and uses.

One of core idea of present embodiment is that adding or click the behavior of using based on the user is that it finds similar application to the preference of using, and according to the promptly historical application of adding or clicking of user's historical preference, recommends similar application to it then.See from the angle of calculating; Exactly all users are come the similarity between the computing application to the preference of certain application as a vector; After the similar application that is applied; Predict that according to the historical preference of user the active user does not also represent the application of preference, the list of application that calculates an ordering is as recommendation.

For making those skilled in the art understand the application better, below through the above-mentioned substep S33-S37 of concrete example description.

Step2：

1). according to the behavior daily record that Mid adds and click application, add up the number of times that Mid adds or click each app;

Application log (nearest 30 days application operating behavior daily record) is perhaps clicked in input data: Mid interpolation;

The form of output data Data1 is: Mid master Appid (app that clicks or add)

Weight3 (the 3rd operation frequency)

For example: the input data are:

27.185.166.230--[20/Aug/2011:10:11:47+0800]″GET/stat.html?type＝open&action＝yingyongdianji&fangshi＝2&Appid＝103352&fenleiid＝10001&from＝0&leixing＝1&style＝fullscreen&uid＝1&pid＝h_home_inst&m＝71ddd8f9f1c84e16438ef109f4b6d77b&modulever＝1.4.0.1041&appver＝1.4.0.1041?HTTP/1.1″2000″-″″Mozilla/4.0(compatible；MSIE?7.0；Windows?NT?6.0；SLCC1；.NET?CLR2.0.50727；Media?Center?PC?5.0；.NET?CLR?3.5.30729；.NET?CLR?3.0.30618)″

Output data Data1 is:

00062ee80feec92758b8be8d3e4b9c16 100113997 2

0006d880a38f687c3fca10e8c4efa227 102019670 2

0007400ba8300fc80b7210a0e66de257 102022805 2

0007bda041324ee53b83f0343daf84d2 102005801 2

00080118032d30ad28a9bce8232d17c8 100013133 3

0008020587ea4fecf08ad58b09ef5904 110195914 2

00082bf0489199360bce8a06693ef3f5 100115004 3

......

Shine upon with the Appid_name table of comparisons, find out the corresponding applicating category of each Appid, the result is following:

00062ee80feec92758b8be8d3e4b9c16 1 100113997 2

0006d880a38f687c3fca10e8c4efa227 1 102019670 2

0007400ba8300fc80b7210a0e66de257 2 102022805 2

0007bda041324ee53b83f0343daf84d2 9 102005801 2

00080118032d30ad28a9bce8232d17c8 6 100013133 3

0008020587ea4fecf08ad58b09ef5904 6 110195914 2

00082bf0489199360bce8a06693ef3f5 4 100115004 3

......

2). in same Mid and feileid, it is clicked the app that perhaps added sort, get preceding 20 Appid, and be included into delegation by clicking or adding number of times.

Input data: Data1 (output data in a last step)

The data structure of output data Data2 is: Mid fenleiid Appid1 weight1 Appid2weight2 Appid3 weight3......

For example: the input data:

00082bf0489199360bce8a06693ef3f5 1 100115004 3

00083ebafe4eb71596f45dfa821f73d5 1 102028904 5

000887c0d3498c7c43cecb566a6333e4 8 102020030 3

0008ce3ae13aa6332794861b275861ad 10?102005157 2

......

Output data Data2 is:

007a2663bac10378bcaa874be36a2d97 1 102006053 2 102028976 1100000913 1

007abe31b117a554df0fefc2a91200c2 2 120042762 11?102023358 2100000568 2 102010364 1 100115004 1

007b3dc8a31a6627a6e2468f789aa078 8 102007826 19?102020628 6102007664 6 102028968 4 100012183 3 100012315 3102043563 2 102022076 2 102000032 2 102031006 2110004672 2 102007377 2 101000009 2 110091072 2100030320 2 102044509 2 102043791 2 102044243 2102044665 2 100040423 2

007bb8043a487ed3a690aa6d461a3c10?10?102019572 3

007d072a6bfe28591f0eb4c5d533784c?18?100000525?31?10000028919?101000053?16?102005903 6 100000625 4 102020030 4102020628 3 120055003 3 100000913 3 102001686 2100034506 2 102005985 2 100000801 2 102044292 2110153628 2 100115575 2 100045261 2 100115650 2100103773 2 100102029 2

3). in a pair of Mid and fenleiid, Appid matches in twos, and note is common to be occurred 1 time, is 1 type with two Appid then, calculates two total degrees that Appid occurs simultaneously, generates frequent 2 collection.

Input data: Data2 (output data in a last step)

The data structure of output data Data3 is: Appid1 Appid2 weight (occurrence number)

For example: the input data

Xx1?1 001?002?003?004......

Xx2?2 001?002?003......

Xx3?15?002?004......

......

The intermediate result file:

001?002?1

001?003?1

001?004?1

002?003?1

002?004?1

003?004?1

001?002?1

002?005?1

002?004?1

Output data Data3 is:

001?002?2

001?003?1

001?004?1

002?003?2

002?004?2

003?004?1

4). calculate frequent 1 collection, note is calculated the number of times that each Appid occurs

The output data in input data: Data1 (the 1st) step)

The data structure of output data Data4 is: Appid weight (occurrence number)

For example:

The input data:

Xx1?001?10

Xx1?002?8

Xx1?003?5

Xx2?002?8

Xx2?003?9

Xx3?003?7

......

Output data Data4 is:

001?1

002?2

003?3

......

5). calculate degree of confidence, and sort by degree of confidence

For example: the number of times that frequent 2 pooled applications A and Application of B occur is N, and the number of times that frequent 1 pooled applications A occurs is M, then with respect to A; The degree of confidence of B is N/M, sorts according to degree of confidence, gets the application of ordering preceding 50; And, merge to delegation with head type of being applied as, structure is:

A?B C?D ......

Input data: Data3 and Data4

The data structure of output data Data5 is: Appid Appid1 weight1 Appid2weight2......Appid50 weight50

For example:

Input data Data3 is:

001?100?20

001?101?18

001?102?16

001?103?14

001?104?12

001?106?10

002?201?50

002?202?30

002?101?10

002?102?5

Input data Data4 is:

001?100

002?200

Output data Data5 is:

001?100?101?102?103?104?106......

002?201?201?101?102?......

6). according to Data1 and Data5, mate, generate recommendation results according to user Appid.

For example:

Data1：

Xx1?1? 001?10

Xx1?2? 002?5

......

Data5：

001?100?101?102?103?104?105?106?107......

002?200?201?202?203?204?205?206?207......

Appid according to indicating underscore matees, and generates intermediate result to be:

Xx1?1?100?101?102?103?104?105?106?107......10

Xx1?2?200?201?202?203?204?205?206?207......5

In with a pair of Mid and fenleiid, sort according to weight and to get preceding 50 and use as follows:

Xx1?1?100?101?102?103?104?105?106?107......200?201?202?203?204?205206?207......

In concrete the realization; Can be further the result of Step1 and Step2 in the above-mentioned example be merged according to Mid and fenleiid; The applicating category of confirming with step 103 filters; Only get the application data of definite applicating category, in each applicating category, get 50 application then at random as recommendation results.If do not have data in certain applicating category, then can adopt arbitrary mechanism to carry out polishing, for example, put into the most popular application, up-to-date application etc.

Certainly; The above-mentioned method of using with the user behavior information matches of searching only is used as example; It also is feasible that those skilled in the art adopt other computing method; For example, the tag along sort through calculating user behavior information and the matching degree of the tag along sort of respective classes application data sets application etc., the application need not this to limit.

Step 105, generate corresponding application file folder, said matched application is put into corresponding application file folder recommend by said applicating category.

Use the application embodiment, category is generated the application file folder, under the respective classes, promptly in the application file folder of corresponding classification, recommend with the application of user behavior information matches to the user, thus the resource that helps saving subscriber equipment.

In concrete the realization, the application file folder for recommending the user can represent in the different split screens of desktop, preferably, can also be according to the height and the width of user's split screen, the number of the application file folder of confirming to recommend in each split screen.Use the application embodiment, the order that represents of said application file folder is that the operation frequency according to the corresponding Main classification label of each applicating category is provided with from high to low, so the application file folder is that matching degree according to user interest represents to the user from high to low; And the application in the application file folder is also sorted by weight, promptly also is that the matching degree according to user interest represents to the user from high to low, thereby operation that can more convenient user makes the user obtain better experience.

In concrete the realization, can unify in the user interface of terminal desktop to show and press from both sides corresponding icon that each icon is represented an application file folder, through the mode of icon conduct with application entrance with a plurality of application files.This patterned exhibition method is very directly perceived for the user, and easy to use and management.For example; The icon of showing the application file folder in the user interface comprises " video "; " novel ", " education " and " recreation " is behind the icon of user's click " video " application file folder; Get into the subwindow of this application file folder, in subwindow, showing has a plurality of application icons such as TV play, film, animation, variety.Can point out the user use through icon as the mode of application entrance to this application; But before the user really selected to use, actual installation should not used corresponding configuration file, like this; Not only can be user-friendly, and before use and exceed and take client resource.

Icon in the user interface can be concentrated by the network side central server and dispose or push, and this has just prevented that rogue program from arbitrarily adding the malice icon in the interface, further improved security.There is the configuration file of central server centralized management can comprise the corresponding reference address of using, present specification, and the unfolding mode of said application, perhaps their any combination.

For example, use for web, the address of web visit is sent to end side by the mode of central server through configuration file, and this has just prevented rogue program the distorting reference address of end side.

And; The network side central server can be through the profile information that upgrades with the mutual acquisition of third party's content server; For example, if the reference address of certain application changes, server can be through the address information after upgrading with the mutual acquisition of content server; And send over through configuration file, stopped to change the opportunity that stays to rogue program because of reference address.

In addition, subscriber equipment can also upgrade the display state of this icon, with further prompting user behind the configuration file of acquisition and the corresponding application of said icon.For example, do not obtain configuration file before, icon can be a black and white, or dark-coloured, and after acquisition, can become colour or light tone.

What also need explain is that the application file clip icon of in the end side user interface, showing can be one or more, can confirm according to different displaying rules.For example, when using an icon, this icon can be used as the unified inlet of the application of a plurality of subordinates or subordinate's icon, and wherein any one is used when obtaining lastest imformation, all can obtain at this inlet icon place to point out.

In a kind of preferred embodiment of the application, can also comprise the steps:

Through setting up the user characteristics storehouse; Then can user behavior information be unified in server end or high in the clouds is handled; In this embodiment; Can in the user characteristics storehouse, work as inferior operation behavior information by recording user, and confirm and to press from both sides and application corresponding to the application file that the user recommends according to the previous operation behavior information in user characteristics storehouse.

Need to prove; For method embodiment, for simple description, so it all is expressed as a series of combination of actions; But those skilled in the art should know; The application does not receive the restriction of described sequence of movement, because according to the application, some step can adopt other orders or carry out simultaneously.Secondly, those skilled in the art also should know, the embodiment described in the instructions all belongs to preferred embodiment, and related action and module might not be that the application is necessary.

With reference to Fig. 2, show a kind of structured flowchart of using automatic recommended device embodiment of the application, specifically can comprise like lower module:

Request receiver module 201 is used to receive the user and obtains request from the application of client submission, and said application is obtained and comprised ID in the request;

Formerly the behavioural information extraction module 202, are used for extracting the existing user behavior information of relative users from the user characteristics storehouse according to said ID, and said user behavior information comprises that the user is directed against the operation information of exemplary application formerly;

Applicating category determination module 203 is used for confirming the applicating category to user's recommendation according to said user behavior information;

Coupling is used acquisition module 204, is used for the application data sets at said applicating category, and being directed against formerly according to the user, the operation information of exemplary application extracts matched application;

Use recommending module 205, be used for generating corresponding application file folder, said matched application is put into corresponding application file folder recommend by said applicating category.

In concrete the realization, the application embodiment can also comprise like lower module:

As the concrete a kind of example used of the application embodiment, said user behavior information also comprises user's local operation behavioural information, and/or, user's online operation behavior information; In this case, said applicating category determination module 203 can comprise following submodule:

In a kind of preferred embodiment of the application, said device can also comprise like lower module:

The application data set generation module is used to generate the application data set of each applicating category:

Said application data set generation module is used acquisition module 204 with coupling and is connected, and specifically can comprise following submodule:

In a kind of preferred embodiment of the application, said coupling is used acquisition module 204 can comprise following submodule:

More preferably, said coupling is used acquisition module 204 and can also be comprised following submodule:

The application embodiment not only can be applied to can also be applied to the applied environment of client-server in the applied environment of single device, perhaps further is applied in the applied environment based on cloud.

Because said device embodiment is basically corresponding to preceding method embodiment, so not detailed part in the description of present embodiment can just not given unnecessary details at this referring to the related description in the previous embodiment.Module, submodule and unit related in the application's device embodiment and the system embodiment can be software, can be hardware, also can be the combination of software and hardware.

The application can be used in numerous general or special purpose computingasystem environment or the configuration.For example: personal computer, server computer, handheld device or portable set, plate equipment, multicomputer system, the system based on microprocessor, set top box, programmable consumer-elcetronics devices, network PC, small-size computer, mainframe computer, comprise DCE of above any system or equipment or the like.

The application can describe in the general context of the computer executable instructions of being carried out by computing machine, for example program module.Usually, program module comprises the routine carrying out particular task or realize particular abstract, program, object, assembly, data structure or the like.Also can in DCE, put into practice the application, in these DCEs, by through communication network connected teleprocessing equipment execute the task.In DCE, program module can be arranged in this locality and the remote computer storage medium that comprises memory device.

At last; Also need to prove; In this article; Relational terms such as first and second grades only is used for an entity or operation are made a distinction with another entity or operation, and not necessarily requires or hint relation or the order that has any this reality between these entities or the operation.And; Term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability; Thereby make and comprise that process, method, article or the equipment of a series of key elements not only comprise those key elements; But also comprise other key elements of clearly not listing, or also be included as this process, method, article or equipment intrinsic key element.Under the situation that do not having much more more restrictions, the key element that limits by statement " comprising ... ", and be not precluded within process, method, article or the equipment that comprises said key element and also have other identical element.

More than a kind of method and automatic recommended device of a kind of application of using automatic recommendation that the application provided carried out detailed introduction; Used concrete example among this paper the application's principle and embodiment are set forth, the explanation of above embodiment just is used to help to understand the application's method and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to the application's thought, the part that on embodiment and range of application, all can change, in sum, this description should not be construed as the restriction to the application.

Claims

1. use the method for recommending automatically for one kind, it is characterized in that, comprising:

2. the method for claim 1 is characterized in that, also comprises:

3. the method for claim 1 is characterized in that, said user behavior information also comprises user's local operation behavioural information, and/or, user's online operation behavior information;

The step of the said applicating category of confirming according to user behavior information to recommend to the user comprises:

4. method as claimed in claim 3 is characterized in that, generates the application data set of certain applicating category through following steps:

Obtain said application quality grading parameters to be recommended;

5. method as claimed in claim 4 is characterized in that, said application data sets at applicating category, and being directed against formerly according to the user, the step of the operation information extraction matched application of exemplary application comprises:

6. method as claimed in claim 5 is characterized in that, said application data sets at applicating category, and being directed against formerly according to the user, the step of the operation information extraction matched application of exemplary application also comprises:

7. use automatic recommended device for one kind, it is characterized in that, comprising:

8. device as claimed in claim 7 is characterized in that, also comprises:

9. device as claimed in claim 7 is characterized in that, said user behavior information also comprises user's local operation behavioural information, and/or, user's online operation behavior information;

Said applicating category determination module comprises:

10. device as claimed in claim 9 is characterized in that, also comprises:

11. device as claimed in claim 10 is characterized in that, said coupling is used acquisition module and is comprised:

12. device as claimed in claim 11 is characterized in that, said coupling is used acquisition module and is also comprised: