CN101105795A - Network behavior based personalized recommendation method and system - Google Patents
Network behavior based personalized recommendation method and system Download PDFInfo
- Publication number
- CN101105795A CN101105795A CNA200610114120XA CN200610114120A CN101105795A CN 101105795 A CN101105795 A CN 101105795A CN A200610114120X A CNA200610114120X A CN A200610114120XA CN 200610114120 A CN200610114120 A CN 200610114120A CN 101105795 A CN101105795 A CN 101105795A
- Authority
- CN
- China
- Prior art keywords
- user
- matching degree
- behavior
- database
- file
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses an individualized recommendation method based network behaviors. The method includes the following steps: (A). Generate the corresponding network behavior log of the user according to the search request sent by the users to the server and the report on the completion of document download, and record the keyword searched by users and downloaded document; (B). Analyze and clean up network behavior log, and modify system knowledge database and user matching database; (C). Find out users with the same interest from the modified user matching database and introduce them to users, and take the documents that users are interested in as the recommended document resource for users. The invention also discloses an individualized recommendation system based on network behaviors, which consists of network behavior log generating module, log analysis handling module, recommendation generating module, knowledge database, and user matching database. The invention has the advantages of high operating efficiency, strong usability of results, easy maintenance and expansion of the system, and capability to trigger users' interest.
Description
Technical field
The present invention relates to a kind ofly provides personalized recommendation service method and system for it on the basis of analysis user network behavior.Particularly, the present invention relates to a kind of by each user's network behavior is analyzed as the keyword and the downloaded files of search, extract the similarity between the user, carry out the personalized recommendation of two aspects as have the good friend of common interest and the method and system of its possible interested various files with it to user's recommendation to the user in view of the above.
Background technology
Follow the develop rapidly of computer networking technology, the internet has become that people obtain, the important platform of exchange of information.Under this background, how to make the user can in the information ocean of Internet, fast, accurately find oneself that need, interested content, become the internet and further developed a difficult problem that needs to be resolved hurrily.Search engine technique is exactly the important tool that solves this difficult problem, and the key word that it provides according to the user returns to the user to the related content that satisfies matched rule, has greatly improved the efficient of information retrieval.But,, can not satisfy the search need of the user of different background, different search purposes and different times well because search engine technique has versatility.Show on the one hand, use search engine, the user must at first extract key word according to the demand of oneself, obviously, the demand whether key word that is proposed can accurately summarize the user badly influences the availability of the return results of search engine, and this has just proposed higher requirement to the ability that the user extracts key word; On the other hand, even the keyword that the user extracts is accurately, search engine also only is simply the result who satisfies coupling to be returned to the user, most content is that the user is unwanted and often have quite among the result, causes the user to have to extract own needed content in the result who returns in a large number.
In order to remedy the deficiency of universal search engine, people have proposed the personalized recommendation technology.The personalized recommendation technology is meant the interest and the behavior of learning the user by Collection and analysis user profile, realizes the purpose of initiatively recommending in view of the above.
At present, the personalized recommendation technology is mainly launched in Web page access field, can be divided into two big classes according to the technology that it adopted: a class is based on the personalized recommendation system of rule; One class is an information filtering system.Wherein, information filtering system can be subdivided into content-based filtering system and collaborative filtering system again.The interested content that rule-based personalized recommendation system was at first read according to the active user, extrapolate the interested content that the user did not also read according to rule, then according to the support (or significance level) of rule, to these content orderings and represent to the user.The extraction of rule is normally found automatically by customization or by digging technology.Rule is many more, and the accuracy of recommendation is high more, but the maintenance of system also becomes difficult more, the reduction that performance also can be rapid.Content-based filtering system is to realize recommending resource to the user by the similarity of calculating user behavior feature and resource characteristic, and the shortcoming of this method is: be difficult to distinguish the quality and the style of resource, and can not find new resources of interest for the user.The system similarity of collaborative filtering system and content-based filtration, but relatively be similarity between the different user, find out several the most similar neighbours users, and come the interest of predictive user by neighbours user, thereby recommend resource to the user, the shortcoming of this method is: the sparse and poor expandability of initial stage system.
Along with the maturation of P2P technology, P2P (abbreviation of Peer to Peer) uses the favor that more and more is subjected to the user, in the P2P network, there is not Centroid that service is provided specially, status between all nodes is an equality, and each node is the requesting party of service, also is the provider of service.P2P has changed the information sharing mode based on the C/S framework in the past, and the user is changed information source initiatively into from passive information acceptor, and the user contributes the resource of this locality out by the P2P system easily.System is along with user's increase, and how the expansion that resource is rapid finds own interested resource or user rapidly in the P2P system, be the difficult problem that the P2P system needs to be resolved hurrily.At present, also be in the starting stage,, have following feature towards the personalized service of P2P system relatively based on the personalized service of Web based on the personalized service of P2P system:
1, the main path of User Recognition file content is the title of file (folder).Generally, the title of file (folder) is brief, does not have contextual aid identification.And identical file may show as different filenames at different user nodes.
2, the content of Tui Jianing not only comprises file resource, also comprises commending friends.The P2P system provides a kind of approach that finds interested content by users interest.
3, the node in the P2P system is unsettled.When user log off (interrupting or the like such as bolt down procedure, shutdown or network), user's resource just becomes inaccessible to other users of system.The Web website is then more stable comparatively speaking, generally is operation in 7 * 24 hours.
Summary of the invention
In analysis-by-synthesis based on the new problem that faces in the pros and cons of the various individuation service systems of Web and the P2P system applies, the personalized recommendation method and the system that the purpose of this invention is to provide a kind of behavior Network Based towards the P2P system, the good friend and the various files of user's interest of common interest can be recommended to have with it for the user based on user's network behavior by this system, for the user provides personalized service.
For achieving the above object, the present invention by the following technical solutions: a kind of personalized recommendation method of behavior Network Based, this method may further comprise the steps:
The first step: the report that searching request of sending to server according to the user and file in download finish generates corresponding this user network behavior daily record, keyword, the downloaded files of recording user search;
Second step: analysis and arrangement is carried out in the network behavior daily record, revise systematic knowledge database and user's matching degree database;
Knowledge data base has been used for since the memory system self-operating, and the diverse network behavior that the user carries out in system comprises user's search behavior and file download behavior;
User's matching degree database is used for the recording user similarity of interest between any two, is to carry out the direct foundation that the user recommends;
The 3rd step: find out friend recommendation according to user's matching degree database of revising and give the user, excavate the common interested file of the similar good friend of these hobbies simultaneously as recommending file resource to recommend the user with same interest hobby.
The invention also discloses a kind of personalized recommendation system of behavior Network Based, it is made of network behavior daily record generation module, log analysis processing module, recommendation generation module, knowledge data base and user's matching degree database;
Described network behavior daily record generation module mainly is to generate corresponding this user network behavior daily record, keyword, the downloaded files of recording user search according to the user to searching request and the file in download conclude report that server sends;
Described network behavior log analysis processing module is carried out analyzing and processing to the daily record that network behavior daily record generation module generates, and revises the systematic knowledge database, and finally revises user's matching degree database;
Described knowledge data base has been used for since the memory system self-operating, and the diverse network behavior that the user carries out in system mainly is that search behavior and file are downloaded behavior; Knowledge data base carries out taxonomic organization according to the MD5 value to network behavior, and according to user ID rank order from small to large;
Described user's matching degree database is used for the recording user similarity of interest between any two, and database is used to write down other users and this user's similarity for each user safeguards a matched data item;
Described recommendation generation module produces according to user's matching degree database and recommends the user and recommend file.
Solution of the present invention is: by the key word and the downloaded files log record of analysis user search, for each user finds out the most similar good friend of several interest, and give the user these friend recommendations; Simultaneity factor is found out the common interested data of this group good friend and is recommended the user by analyzing the similar good friend's of interest data bank.The inventive method need be safeguarded two databases, is respectively knowledge data base and user's matching degree database.Knowledge data base is a key word with the keyword and the downloaded files of search, has write down the user who searched for this keyword or downloaded this document in system's operational process.When the user used a certain keyword search for the first time or downloads a certain file for the first time, knowledge data base was learnt this behavior of user.Knowledge data base is the core data of system, and it has write down all users' network behavior in the system, is the basis that system carries out friend recommendation and file recommendation.User's matching degree database is used for the number of times that the identical network behavior takes place recording user between any two, as uses identical keyword search or download identical file.
When will carry out friend recommendation, system retrieves several other users the most close with recommended user interest as recommended from user's matching degree database.The degree that interest is close between the user makes a decision according to its matching degree, and matching degree is high more, and the interest between the user is close more.The user who searches for identical keyword or download same file has identical objects, and the matching degree between the user remembers from 0, and when a same keyword search taking place or download identical file, just the matching degree to respective user adds 1.Find out with after certain particular user has identical interested some other users according to matching degree, from these other users, find out some file resources that they own together again and recommend this particular user.
The present invention has the operational efficiency height, availability is strong as a result, system is easy to maintenance and expansion, can excite the characteristics of the new interest of user.
1, operational efficiency height
This method only relates to simple data processing, mainly is to merge and ordering simple the classification, does not have complicated data processing, and computation complexity is nlog
nTraditional relatively personalized recommendation system, operational efficiency is greatly improved.
2, availability is stronger as a result
Existing personalized recommendation system based on web is to infer its point of interest by the content of user's browsing page, and then produces the content of recommending.In fact, the keyword of search and downloaded files can both characterize user's point of interest to a certain extent, and since the expense of file in download greater than the expense of opening the page, the keyword that downloaded files compares more helps characterizing user's point of interest.Therefore, only infer that from searching key word user's point of interest has its one-sidedness.Extract based on the user characteristics of keyword like this and resource characteristic extracts and the recommendation results that produces of coupling is also just unsatisfactory naturally thus.Recommend method of the present invention is from basic existence general knowledge, be the center with certain targeted customer successively, other users that have a common network behavior (keyword and downloaded files and the shared network behaviors such as file that comprise search) with this user are classified as the customer group with common interest hobby, then the common search of other users among the group or downloaded but file that this targeted customer does not also search for or downloaded is recommended him.The rationality of the inventive method is: the first, and it meets the logic of daily life, has the specific crowd of similar behavior generally also to have common or similar hobby and demand; The second, it only carries out cluster according to user behavior to the user, does not need user characteristics and resource characteristic are extracted, and has avoided the one-sidedness based on the user characteristics extraction of keyword.Thus, to produce the recommendation results availability also just stronger naturally in the present invention.
3, be easy to maintenance and expansion
The relation of relation between the user and user and resource all is a highly sparse matrix.Two databases of the inventive method design are only safeguarded the matrix nonzero element, and the data volume of maintenance is little, and the expense of calculating reduces greatly; When carrying out resource recommendation, only relate to the analysis of very little a part of resource; The design of simultaneity factor had both been supported to adopt by different level and had been handled, and also supported distributed treatment, had good maintainability and extensibility.
4, can excite user's new interest
The present invention is based on user's network behavior analysis, other users that have a common network behavior with certain targeted customer are classified as the customer group with common interest hobby, then other users among the group are searched for jointly, download or shared but file that this specific user did not also search for, downloads or shared is recommended him.The result that this recommend method produces not only has higher availability, and many times or even user institute beyond thought.The point of interest that can excite the user to hide not only, but also have intelligentized learning functionality.
Description of drawings
Fig. 1 is that the personalized recommendation system system that the present invention is based on network behavior constitutes synoptic diagram;
Fig. 2 is the personalized recommendation method overview flow chart that the present invention is based on network behavior;
Fig. 3 is the network behavior log analysis flow chart of steps that the present invention is based in the personalized recommendation method of network behavior;
Fig. 4 revises the knowledge data base flow chart of steps for the present invention;
Fig. 5 revises user's matching degree database steps process flow diagram for the present invention;
Fig. 6 recommends to generate flow chart of steps for the present invention;
Fig. 7 is the personalized recommendation system processing flow chart of the distributed behavior Network Based of the present invention.
Embodiment
Fig. 1 is that the personalized recommendation system system of behavior Network Based disclosed by the invention constitutes synoptic diagram.As shown in the figure, the personalized recommendation system of behavior Network Based disclosed by the invention mainly is made of network behavior daily record generation module, log analysis processing module, recommendation generation module, knowledge data base and user's matching degree database; The log analysis processing module comprises daily record order module, knowledge data base modified module again, appends matching degree list ordering module and user's matching degree database update module; Recommend generation module to comprise again and generate the commending friends submodule and generate recommendation file submodule.
Because the present invention do not require system real time, the triggering of system generally is arranged in the moment of server free time every day start-up time and carries out.Can consider to select the regularly mechanism of triggering, also can select the mechanism of manual triggers.
To be introduced network behavior daily record generation module, log analysis processing module, knowledge data base, user's matching degree database, recommendation generation module respectively below.
1, network behavior daily record generation module
Because the present invention mainly proposes at the P2P shared file system, the user mainly is that search and file are downloaded in the behavior of internal system, so grid behavior daily record generation module mainly is to generate corresponding this user network behavior daily record, the keyword of recording user search, downloaded files etc. according to the user to searching request and the file in download conclude report that server sends.MD5 value (full name of MD5 is Message-Digest Algorithm 5, and md5-challenge promptly is transformed into the byte serial of a random length the big integer of one fixed length) has been introduced in the management of log pattern for convenience in the log record.The user of P2P is the legal registered user of system, and in the process of registration, system distributes an ID that system is unique for each user.The all corresponding MD5 value of each keyword or each file.The form of journal file is as follows:
| Clauses and subclauses | Content |
| Filename | ssn.log |
| File layout | md5?userid?timeStamp?opContent ... |
| Format description | OpConeten:opCode[arg0] [arg1] [... ] 1. share opCode:1 opCode fileName fileSize 2. download opCode:2 opCode fileName filesize 3. search opCode:3 opCode keyword |
| Example | Asdfkjalsdfjlasdjflasjds 1,000 1,384,234,234 1 red light .rm 10000000000 ajsdjflasjfdasldfjalsdjfl, 1,003 1,393,243,434 2 shuttlecock .txt 19934923423 sfsfjasldfjlasfdjlasjdflas, 1,000 1,394,333,333 3 artificial intelligence |
Network behavior daily record generation module is exactly the network behavior each time by user ID and the MD5 value record user corresponding with each behavior, generates daily record.
2, network behavior log analysis processing module
Network behavior log analysis processing module is the core processing module of system.This module is carried out analyzing and processing to the daily record that network behavior daily record generation module generates, and revises the systematic knowledge database, and finally revises user's matching degree database, for recommending to get ready.This module is a module the most complicated in the system, and its performance has determined the performance of system.For the performance of optimization system, improve the processing power of system, this module includes daily record order module (being the daily record pretreatment module again), knowledge data base modified module, appends matching degree list ordering module and user's matching degree database update module.
Along with the increase of system user, the increase that the each log record number to be processed of system also can be rapid, systematic knowledge database and the also dilatation accordingly of system user matching degree database.For the processing power of optimization system, improve the extendability of system, this module is supported the distributed treatment pattern.
The keyword of each search or each downloaded files all have corresponding M D5 value corresponding with it, and each MD5 value all has corresponding record corresponding with it.According to common processing mode, log record of every processing will be distinguished database of reading and writing, promptly carry out a disk I operation.In order to reduce the disk I operation, introduce the daily record order module.The daily record order module sorts to the log record that the daily record generation module produces according to the MD5 value, make log record be aggregated in one with identical MD5 value, like this, only need carry out one time the database read write operation, significantly reduce the number of times of database access the record of identical MD5 value.
The knowledge data base modified module is mainly finished the modification of knowledge data base, and the tabulation of user's matching degree is appended in generation.
Append matching degree list ordering module and mainly be to append the tabulation of user's matching degree according to targeted customer ID to matching degree tabulation sort.
User's matching degree database update module mainly is to revise user's matching degree database.
3, knowledge data base and user's matching degree database
Knowledge data base has been mainly used in since the memory system self-operating, and the diverse network behavior that the user carries out in system mainly is that search behavior and file are downloaded behavior.Simultaneously, knowledge data base is the basis that user's matching degree database produces.Knowledge data base is the critical data storehouse of system, if the network behavior daily record is lost, knowledge data base can't recover; Even log record has preservation, recover knowledge data base from the history log of system also will be a complexity, time-consuming procedure, therefore be necessary knowledge data base is carried out redundancy backup.
The organization and administration of data for convenience, knowledge data base carries out taxonomic organization according to the MD5 value to network behavior.The corresponding data item of each MD5 value is because the generating function of MD5 value not is one to one, so not only keyword or file of possible correspondence in the data item of each MD5 value correspondence.Here adopt XML to organize the network behavior with same MD5 value, concrete formal definition is as follows:
| Clauses and subclauses | Content |
| Filename | (md5).xml |
| File layout | <record?md5＝xxxx?lastModifyTime＝xxxx> <download> <item?size＝xxx> <userlist>ID1&ID2</userlist> </item> </download> <search> <item?word＝””> <userlist>ID1&ID2</userlist> </item> </search> </record> |
| Example | <record md5=fsdfflsdlsdfsadfsdfsdfds lastModifyTime=1 11111111><download><item size=13456775><userlist>1000 ﹠ 1001</userlist></item></download><search><item word=" computer "><userlist>1006 ﹠ 1008</userlist></item></search></record> |
Wherein the user list of each keyword or file correspondence uses ﹠amp; User ID is separated mutually, and required the satisfied order from small to large of ordering of user ID, the conflict inspection of convenient follow-up insertion.
User's matching degree database is used for the recording user similarity of interest between any two, is to carry out the direct foundation that the user recommends.This database root it is reported that knowing user's matching degree tabulation of appending that the database update module produces safeguards.Because the similarity matrix between the user is a height sparse matrix, database adopts the mode of only storing nonzero element to compress storage space for this reason.The generation of recommending for convenience, database is used to write down other users and this user's similarity (only writing down the element of similarity non-0) for each user safeguards a matched data item.The matched data item adopts the format organization of XML, and concrete form is as follows:
| Clauses and subclauses | Content |
| Filename | (userid)Match.xml |
| File layout | <match?id＝xxx?lastModifyTime＝xxx> <user?id＝xxx>xxx</user> <user?id＝xxx>xxx</user> <user?id＝xxx>xxx</user> </match> |
| Example | <match?id＝1001lastModifyTime＝12343221> <user?id＝1005>1</user> <user?id＝1006>2</user> <user?id＝1007>1</user> </match> |
Wherein user's ordering is ascending according to user ID, makes things convenient for subsequent modification and insertion.When the matched data item is modified, last modification time (lastModifyTime) also can be refreshed accordingly, recommends generation module to judge whether to regenerate the recommendation user easily according to last this field of modification time (lastModifyTime).
4, recommend generation module
Recommending generation module is the module that system produces the recommendation user and recommends file, and this module generates submodule by friend recommendation and file recommends the generation submodule to form.
Friend recommendation generates submodule and mainly is responsible for generating the good friend who recommends to the user.This module retrieves the good friend who recommends the user from user's matching degree database.At first to sorting according to matching degree with the user who recommends target to have the same interest hobby, several users that select matching degree value maximum then are as the user who recommends.
File is recommended to generate submodule and mainly is responsible for generating the file of recommending to the user.The logical relation of this module institute foundation is: having the common interested file resource of other users of similar network behavior with certain targeted customer, also can be that this targeted customer institute is interested.File is recommended to generate submodule and is generated the good friend that submodule produces according to friend recommendation, analyzes their common interested file, and recommends the targeted customer.In the system of P2P, each user can share or file in download.This module is shared or downloaded files filters out their common interested some files and recommends the user from commending friends.
Fig. 2 is the personalized recommendation method overview flow chart that the present invention is based on network behavior.As shown in the figure, the personalized recommendation method of behavior Network Based provided by the invention may further comprise the steps:
The first step: the report that searching request that grid behavior daily record generation module sends to server according to the user and file in download finish generates corresponding this user network behavior daily record, keyword, the downloaded files of recording user search.
System distributes an ID that system is unique for each user, simultaneously, system is that each keyword, each file all are provided with a MD5 value accordingly, the system journal generation module is according to user's network behavior, with ID and MD5 value record user's network behavior each time, generate user's network behavior daily record.The form of journal file is as follows:
| Clauses and subclauses | Content |
| Filename | ssn.log |
| File layout | md5?userid?timeStamp?opContent ?... |
| Format description | OpConeten:opCode[arg0] [arg1] [...] 1. share opCode:1 opCode fileName fileSize 2. download opCode:2 opCode fileName filesize 3. search opCode:3 opCode keyword |
| Example | Asdfkjalsdfjlasdjflasjds 1,000 1,384,234,234 1 red light .rm 10000000000 ajsdjflasjfdasldfjalsdjfl, 1,003 1,393,243,434 2 shuttlecock .txt 19934923423 sfsfjasldfjlasfdjlasjdflas, 1,000 1,394,333,333 3 artificial intelligence |
Second step: by grid behavior log analysis processing module analysis and arrangement is carried out in daily record, and revise corresponding knowledge data base and user's matching degree database.
As shown in Figure 3, analysis and arrangement is carried out in daily record, and the concrete steps of revising corresponding knowledge data base and user's matching degree database are by grid behavior log analysis processing module:
1, the daily record order module in the network behavior log processing module sorts to log record according to the MD5 value, makes the log record with identical MD5 value be aggregated in one.
2, by the knowledge data base modified module knowledge data base is made amendment, and the matching degree tabulation is appended in generation.
Knowledge data base is to be used for since the memory system self-operating, and the diverse network behavior that the user carries out in system mainly is that search behavior and file are downloaded behavior.Knowledge data base carries out taxonomic organization according to the MD5 value to network behavior, and according to user ID sequential storing data from small to large, the conflict inspection of convenient follow-up insertion.
The knowledge data base modified module is revised knowledge data base, generate append the matching degree tabulation detailed process as shown in Figure 4,
(1), at first, from putting into the waiting task pond through reading log record the network behavior journal file after the arrangement of daily record order module with identical MD5 value;
(2), then, confirm whether being recorded in the knowledge data base of corresponding MD5 value exists,, if there is no, then in knowledge data base, create this record if exist then read it;
(3), then, take out daily record one by one and handle from task pool, unwritten log record in the knowledge data base is carried out record, the registered user generates and appends the tabulation of user's matching degree, and the daily record of having write down is filtered, and directly ignores this daily record; Until all log records of finishing dealing with and having identical MD5 value.
Because used keyword search of a knowledge data base recording user and downloaded files do not write down the number of times of use, so, be necessary the daily record that repeats is filtered.When data item is made amendment, taken in if find this network behavior, then directly ignore this daily record; If also there is not record, then write down it, and produce and to append the operation of user's matching degree, write and append the tabulation of user's matching degree.
Append matching degree tabulation and be used to instruct retouching operation user's matching degree database.Its form is as follows:
| Clauses and subclauses | Content |
| Filename | (host).mml |
| File layout | destUserID?userID?value destUserID?userID?value |
| Example | 1000?1001?1 1002?1001?3 |
Wherein " 1,000 1,001 1 " expression ID appends 1 for user 1001 matching degree in 1000 user's the matching degree data item.The matching degree of appending can be respectively given according to the different operating type, our simple same keyword of given every search or download a same file here, and matching degree adds 1.
3, appending matching degree list ordering module sorts to appending the tabulation of user's matching degree
In user's matching degree database, the record of matching degree writes down this user and other users' matching degree respectively with each user-center.Knowledge data base is similar with revising, and revises a user's matching degree data item at every turn, all must carry out a database read operation and a write operation.In order to improve database access efficient, make it possible to same user's data item is revised once concentrated carrying out, network behavior log analysis processing module also comprises appends matching degree list ordering module.
This append matching degree list ordering module to append user's matching degree tabulation according to targeted customer ID to matching degree tabulation sort.
4, user's matching degree database update module is revised user's matching degree database
As shown in Figure 5, the concrete steps of modification user matching degree database comprise:
(1), at first, appending after process is appended the arrangement of matching degree list ordering module read the operation of appending with targeted customer and put into the waiting task pond the matching degree tabulation;
(2), then, confirm that this targeted customer is recorded in user's matching degree database whether to exist, if exist then refresh the last refresh time of this user; If there is no, then in user's matching degree database, create record, and refresh the last refresh time of this user;
(3), then, the matching degree of taking out this user from task pool is one by one appended operation and is handled, to unwritten corresponding matching degree relation in user's matching degree database, register, to recording the modification matching degree of corresponding matching degree; Append operation until handle whole matching degrees of this user, finish.
Revise user's matching degree database and revise the different of knowledge data base and be: the first, user's matching degree is just revised in the modification of user's matching degree database, the output of coming to nothing, and the modification knowledge data base can generate one at last and appends user's matching degree and tabulate; The second, the modification of user's matching degree database not only will be revised and this subscriber-related matching degree, refreshes last modification time but also will finish.
The 3rd step: recommend generation module to find out friend recommendation by the grid behavior and give the user with same interest hobby according to user's matching degree database of revising, excavate the common interested file of the similar good friend of these hobbies simultaneously as recommending file resource to recommend the user, as shown in Figure 6.
Friend recommendation generates submodule and mainly is responsible for generating the good friend who recommends to the user.At first to sorting according to matching degree with the user who recommends target to have the same interest hobby, several users that pick out matching degree value maximum then from user's matching degree database are as the user who recommends for this module.
File recommends the good friend who generates the generation of submodule analysis friend recommendation generation submodule to share or downloaded files, therefrom filters out their common interested file, recommends the targeted customer.
The personalized recommendation method of behavior Network Based provided by the invention and system can realize repertoire on the main frame of unit framework, also can realize repertoire on distributed multiple host.
So-called unit framework is meant the repertoire of realizing the personalized recommendation service by a bare machine.Under the unit framework, the logical operation of system only relates to local logical operation.This system can move on as various different operating system platforms such as windows, linux, unix, can select various programming languages such as C, C++, Java, Dephi, VB.The database that system relates to can be selected Oracle, business databases such as SQL Server, DB, also can select databases such as free MySQL, msSQL, can also use file system database.
Along with the increase of system user, pending daily record also can increase, and the data of database amount also can increase rapidly.The tupe of unit structure can't in time be handled daily record and generate recommendation results, for adaptive faculty, elastic load ability and the extensibility that improves system, therefore, the present invention has introduced the distributed collaborative treatment mechanism in system design, as shown in Figure 6.Fig. 6 has described N platform main frame collaborative process process flow diagram, and every machine is responsible for safeguarding the knowledge of a MD5 section, safeguards user's matching degree data of a user ID section simultaneously.The collection of network log also is to adopt the mode of distributed collection to produce.
At first, every machine is sorted out ordering oneself collecting the network behavior daily record that generates, and the segmentation according to the MD5 value is pushed to corresponding daily record on the processor of being responsible for this MD5 section then.This propelling movement both can adopt self-defining procotol to push, and also can adopt prior protocols, as FTP etc.Target processor carries out sort merge to daily record once more receiving after every other processor pushes the daily record of coming, and carries out the processing and the generation of daily record subsequently and appends the matching degree tabulation.What generate appends the matching degree tabulation at first in the arrangement of sorting of this machine, then according to the segmentation of user ID appending accordingly on the processor that operating list is pushed to responsible this ID section, the mode of propelling movement is as daily record.Target processor pushes appending after the matching degree tabulation of coming receiving other all main frames, merges arrangement to appending the matching degree tabulation once more, makes amendment according to list for user matching degree database then.Distributed in order better to support, the segmentation of MD5 value and the segmentation of user ID adopt the mode of configuration file to be described.Like this, when increasing or reduce processor, only need make amendment to configuration file gets final product.
The above only is for the demonstration case that illustrates that principle of the present invention and personalized recommendation service produce, and any equivalent transformation based on the principle of the invention and demonstration case all belongs within the protection domain of the present invention.
Claims (14)
1. the personalized recommendation method of a behavior Network Based, this method may further comprise the steps:
The first step: the report that searching request of sending to server according to the user and file in download finish generates corresponding this user network behavior daily record, keyword, the downloaded files of recording user search;
Second step: analysis and arrangement is carried out in the network behavior daily record, revise systematic knowledge database and user's matching degree database;
Knowledge data base has been used for since the memory system self-operating, and the diverse network behavior that the user carries out in system comprises user's search behavior and file download behavior;
User's matching degree database is used for the recording user similarity of interest between any two, is to carry out the direct foundation that the user recommends;
The 3rd step: find out friend recommendation according to user's matching degree database of revising and give the user, excavate the common interested file of the similar good friend of these hobbies simultaneously as recommending file resource to recommend the user with same interest hobby.
2. the personalized recommendation method of behavior Network Based according to claim 1, it is characterized in that: the described first step: the concrete grammar that generates the network behavior daily record is: system distributes an ID that system is unique for each user, simultaneously, system is that each keyword, each file all are provided with a MD5 value accordingly; System is according to user's network behavior, with ID and MD5 value record user's network behavior each time, as user network behavior daily record.
3. the personalized recommendation method of behavior Network Based according to claim 1 and 2, it is characterized in that: described second step: analysis and arrangement is carried out in the network behavior daily record, revise systematic knowledge database and user's matching degree database and further may further comprise the steps:
A, log record is sorted, make log record be aggregated in one with identical MD5 value according to the MD5 value;
B, knowledge data base is made amendment, and generate and append the matching degree tabulation;
(1), at first, from putting into the waiting task pond through reading log record the network behavior journal file after the ordering arrangement with identical MD5 value;
(2), then, confirm whether being recorded in the knowledge data base of corresponding MD5 value exists,, if there is no, then in knowledge data base, create this record if exist then read it;
(3), then, take out daily record one by one and handle from task pool, unwritten log record in the knowledge data base is carried out record, the registered user generates and appends the tabulation of user's matching degree, and the daily record of having write down is filtered, and directly ignores this daily record; Until all log records of finishing dealing with and having identical MD5 value;
Append matching degree tabulation and be used to instruct retouching operation to user's matching degree database, its form is as follows:
Wherein " destUserID userID value " expression ID appends value numerical value for the matching degree of user userID in user's the matching degree data item of destUserID;
C, sort to appending user's matching degree tabulation
To append user's matching degree tabulation according to targeted customer ID to matching degree tabulation sort;
D, modification user matching degree database
Concrete steps are:
(1), puts into the waiting task pond from reading the operation of appending the matching degree tabulation with targeted customer through appending after the ordering arrangement;
(2), then, confirm that this targeted customer is recorded in user's matching degree database whether to exist, if exist then refresh the last refresh time of this user; If there is no, then in user's matching degree database, create record, and refresh the last refresh time of this user;
(3), then, the matching degree of taking out this user from task pool is one by one appended operation and is handled, to unwritten corresponding matching degree relation in user's matching degree database, register, to recording the modification matching degree of corresponding matching degree; Append operation until handle whole matching degrees of this user, finish.
4. the personalized recommendation method of behavior Network Based according to claim 3, it is characterized in that: described the 3rd step: find out friend recommendation according to user's matching degree database of revising and give the user, excavate the common interested file of the similar good friend of these hobbies simultaneously and be as the concrete grammar of recommending file resource to recommend the user with same interest hobby:
At first, to sorting according to matching degree with the user who recommends target to have the same interest hobby, several users that pick out matching degree value maximum then from user's matching degree database are as the user who recommends;
Analysis user and commending friends are shared or downloaded files, therefrom filter out their common interested file as recommending file, recommend the targeted customer.
5. the personalized recommendation method of a behavior Network Based, it is characterized in that: this method is finished by N platform main frame collaborative process, and every main frame is responsible for safeguarding the knowledge of a MD5 section, safeguards user's matching degree data of a user ID section simultaneously; The collection of network log also is to adopt the mode of distributed collection to produce; It may further comprise the steps:
The report that searching request that A, every main frame send to server according to the user and file in download finish generates corresponding this user network behavior daily record;
B, every machine are sorted out ordering oneself collecting the network behavior daily record that generates according to the MD5 value; Then, the segmentation according to the MD5 value is pushed to corresponding daily record on the processor of being responsible for this MD5 section;
C, target processor carry out sort merge to daily record once more receiving after every other processor pushes its MD5 section daily record of being responsible for of coming;
The matching degree tabulation is appended in D, modification knowledge data base and generation;
E, at first appends the matching degree tabulation arrangement of sorting to what generate on this machine; Then, the segmentation according to user ID is pushed on the processor of being responsible for this ID section appending the matching degree tabulation accordingly;
F, target processor push appending after the matching degree tabulation of coming receiving other all main frames, merge arrangement to appending the matching degree tabulation once more;
G, make amendment according to list for user matching degree database;
H, generate commending friends and recommend file, be transferred to the user according to user's matching degree database.
6. the personalized recommendation method of behavior Network Based according to claim 5, it is characterized in that: described steps A: the concrete grammar that generates the network behavior daily record is: system distributes an ID that system is unique for each user, simultaneously, system is that each keyword, each file all are provided with a MD5 value accordingly; System is according to user's network behavior, with ID and MD5 value record user's network behavior each time, as user network behavior daily record.
7. the personalized recommendation method of behavior Network Based according to claim 6 is characterized in that: described step D: revise knowledge data base and generate the concrete grammar that appends the matching degree tabulation and be;
(1), at first, from putting into the waiting task pond through reading log record the network behavior journal file after the ordering arrangement with identical MD5 value;
(2), then, confirm whether being recorded in the knowledge data base of corresponding MD5 value exists,, if there is no, then in knowledge data base, create this record if exist then read it;
(3), then, take out daily record one by one and handle from task pool, unwritten log record in the knowledge data base is carried out record, the registered user generates and appends the tabulation of user's matching degree, and the daily record of having write down is filtered, and directly ignores this daily record; Until all log records of finishing dealing with and having identical MD5 value.
8. the personalized recommendation method of behavior Network Based according to claim 7 is characterized in that: among described step e, the F: to appending the method that the tabulation of user's matching degree sorts be: according to targeted customer ID to matching degree tabulation sort.
9. the personalized recommendation method of behavior Network Based according to claim 8, it is characterized in that: described step G: the concrete grammar of making amendment according to list for user matching degree database is:
(1), puts into the waiting task pond from reading the operation of appending the matching degree tabulation with targeted customer through appending after the ordering arrangement;
(2), then, confirm that this targeted customer is recorded in user's matching degree database whether to exist, if exist then refresh the last refresh time of this user; If there is no, then in user's matching degree database, create record, and refresh the last refresh time of this user;
(3), then, the matching degree of taking out this user from task pool is one by one appended operation and is handled, to unwritten corresponding matching degree relation in user's matching degree database, register, to recording the modification matching degree of corresponding matching degree; Append operation until handle whole matching degrees of this user, finish.
10. the personalized recommendation method of behavior Network Based according to claim 9 is characterized in that: described step H: generate commending friends and recommend file according to user's matching degree database, the concrete grammar that is transferred to the user is:
(1), to sorting according to matching degree with the user who recommends target to have the same interest hobby, several users that pick out matching degree value maximum then from user's matching degree database are as the user who recommends;
(2), analysis user and commending friends share or downloaded files, therefrom filters out their common interested file, recommends the targeted customer.
11. the personalized recommendation system of a behavior Network Based is characterized in that: it is made of network behavior daily record generation module, log analysis processing module, recommendation generation module, knowledge data base and user's matching degree database;
Described network behavior daily record generation module mainly is to generate corresponding this user network behavior daily record, keyword, the downloaded files of recording user search according to the user to searching request and the file in download conclude report that server sends;
Described network behavior log analysis processing module is carried out analyzing and processing to the daily record that network behavior daily record generation module generates, and revises the systematic knowledge database, and finally revises user's matching degree database;
Described knowledge data base has been used for since the memory system self-operating, and the diverse network behavior that the user carries out in system mainly is that search behavior and file are downloaded behavior; Knowledge data base carries out taxonomic organization according to the MD5 value to network behavior, and according to user ID rank order from small to large;
Described user's matching degree database is used for the recording user similarity of interest between any two, and database is used to write down other users and this user's similarity for each user safeguards a matched data item;
Described recommendation generation module produces according to user's matching degree database and recommends the user and recommend file.
12. the personalized recommendation system of behavior Network Based according to claim 11, it is characterized in that: described network behavior daily record generation module utilizes user ID and the MD5 value record user's corresponding with each network behavior network behavior each time, generates daily record.
13. the personalized recommendation system according to claim 11 or 12 described behaviors Network Based is characterized in that: described network behavior log analysis processing module comprises daily record order module, knowledge data base modified module, appends matching degree list ordering module and user's matching degree database update module;
The daily record order module sorts to the log record that the daily record generation module produces according to the MD5 value, makes the log record with identical MD5 value be aggregated in one;
The knowledge data base modified module is mainly finished the modification of knowledge data base, and the tabulation of user's matching degree is appended in generation;
Append matching degree list ordering module and mainly be to append the tabulation of user's matching degree according to? tabulation is sorted to matching degree;
User's matching degree database update module mainly is to revise user's matching degree database.
14. the personalized recommendation system of behavior Network Based according to claim 13 is characterized in that: described recommendation generation module generates submodule by friend recommendation and file recommends the generation submodule to form;
Friend recommendation generates submodule and is responsible for generating the good friend who recommends to the user, at first to sorting according to matching degree with the user who recommends target to have the same interest hobby, several users that select matching degree value maximum then are as the user who recommends for this module;
File recommend to generate submodule and is responsible for generating the file of recommending to the user, this module from commending friends share or downloaded files filter out they jointly interested some files recommend the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200610114120XA CN100530185C (en) | 2006-10-27 | 2006-10-27 | Network behavior based personalized recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200610114120XA CN100530185C (en) | 2006-10-27 | 2006-10-27 | Network behavior based personalized recommendation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101105795A true CN101105795A (en) | 2008-01-16 |
| CN100530185C CN100530185C (en) | 2009-08-19 |
Family
ID=38999693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB200610114120XA Expired - Fee Related CN100530185C (en) | 2006-10-27 | 2006-10-27 | Network behavior based personalized recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100530185C (en) |
Cited By (83)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101188521B (en) * | 2007-12-05 | 2010-07-14 | 北京金山软件有限公司 | A method for digging user behavior data and website server |
| CN101909018A (en) * | 2009-06-05 | 2010-12-08 | 潘晓丰 | Method and system for returning to instant messaging group according to webpage browsed by user |
| CN101916286A (en) * | 2010-08-23 | 2010-12-15 | 宇龙计算机通信科技(深圳)有限公司 | Information recommendation method and system |
| CN101957845A (en) * | 2010-09-17 | 2011-01-26 | 百度在线网络技术（北京）有限公司 | On-line application system and implementation method thereof |
| CN101984420A (en) * | 2010-09-03 | 2011-03-09 | 百度在线网络技术（北京）有限公司 | Method and equipment for searching pictures based on word segmentation processing |
| CN102065454A (en) * | 2010-11-04 | 2011-05-18 | 北京邮电大学 | Context-based method for estimating amount of information |
| CN102130896A (en) * | 2010-01-14 | 2011-07-20 | 腾讯科技（深圳）有限公司 | Method and system for correlating network applications |
| WO2011088723A1 (en) * | 2010-01-20 | 2011-07-28 | 腾讯科技（深圳）有限公司 | Method and system for recommending friends in social networking service (sns) community |
| CN102254039A (en) * | 2011-08-11 | 2011-11-23 | 武汉安问科技发展有限责任公司 | Searching engine-based network searching method |
| CN102279851A (en) * | 2010-06-12 | 2011-12-14 | 阿里巴巴集团控股有限公司 | Intelligent navigation method, device and system |
| CN102456019A (en) * | 2010-10-18 | 2012-05-16 | 腾讯科技（深圳）有限公司 | Retrieval method and device |
| CN101616292B (en) * | 2008-06-25 | 2012-05-23 | 索尼株式会社 | Information processing apparatus, information processing method, and information processing system |
| WO2012065478A1 (en) * | 2010-11-18 | 2012-05-24 | 中兴通讯股份有限公司 | Method and apparatus for intelligent data push |
| CN101751448B (en) * | 2009-07-22 | 2012-06-27 | 中国科学院自动化研究所 | Commendation method of personalized resource information based on scene information |
| WO2012088720A1 (en) * | 2010-12-31 | 2012-07-05 | Yahoo! Inc. | Behavioral targeted social recommendations |
| CN102663073A (en) * | 2012-03-31 | 2012-09-12 | 奇智软件（北京）有限公司 | Method and system for recommending based on downloaded files |
| CN102687166A (en) * | 2009-12-31 | 2012-09-19 | 诺基亚公司 | Methods and apparatuses for user interest modeling |
| CN102685224A (en) * | 2012-04-28 | 2012-09-19 | 华为技术有限公司 | User behavior analysis method, related equipment and system |
| WO2012122754A1 (en) * | 2011-03-15 | 2012-09-20 | 苏州阔地网络科技有限公司 | Method and apparatus for filtering dynamic information |
| CN102693223A (en) * | 2011-03-21 | 2012-09-26 | 潘燕辉 | Search method |
| CN101923545B (en) * | 2009-06-15 | 2012-10-10 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN102724290A (en) * | 2012-05-23 | 2012-10-10 | 华为技术有限公司 | Method, device and system for getting target customer group |
| CN102750334A (en) * | 2012-06-01 | 2012-10-24 | 北京市农林科学院农业科技信息研究所 | Agricultural information accurate propelling method based on data mining (DM) |
| CN102855309A (en) * | 2012-08-21 | 2013-01-02 | 亿赞普(北京)科技有限公司 | Information recommendation method and device based on user behavior associated analysis |
| CN102870114A (en) * | 2010-04-28 | 2013-01-09 | Jvc建伍株式会社 | Item selection device, item selection method, and item selection-use program |
| CN102867001A (en) * | 2011-07-05 | 2013-01-09 | 腾讯科技（北京）有限公司 | Method and device for acquiring information |
| CN102982042A (en) * | 2011-09-07 | 2013-03-20 | 中国移动通信集团公司 | Personalization content recommendation method and platform and system |
| CN103064869A (en) * | 2012-09-05 | 2013-04-24 | 常州嘴馋了信息科技有限公司 | User profile editing system |
| CN103077229A (en) * | 2013-01-04 | 2013-05-01 | 广东一一五科技有限公司 | Method and system for matching user groups |
| CN103106225A (en) * | 2011-11-10 | 2013-05-15 | 北京千橡网景科技发展有限公司 | Method and device for confirming preference for user to select friends in social network |
| CN103116589A (en) * | 2011-11-17 | 2013-05-22 | 腾讯科技（深圳）有限公司 | Method and device of sending recommendation information |
| CN103136269A (en) * | 2011-12-02 | 2013-06-05 | 潘燕辉 | Method of positioning crowds according to program |
| CN103166918A (en) * | 2011-12-12 | 2013-06-19 | 阿里巴巴集团控股有限公司 | Data recommendation method and device |
| CN103218366A (en) * | 2012-01-20 | 2013-07-24 | 腾讯科技（深圳）有限公司 | Method and system for recommending download resources |
| WO2013139239A1 (en) * | 2012-03-21 | 2013-09-26 | Tencent Technology (Shenzhen) Company Limited | Method for recommending users in social network and the system thereof |
| CN103368994A (en) * | 2012-03-28 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Specific data push method and system, and retrieval server |
| CN103425650A (en) * | 2012-05-15 | 2013-12-04 | 腾讯科技（深圳）有限公司 | Recommendation searching method and recommendation searching system |
| CN103577486A (en) * | 2012-08-02 | 2014-02-12 | 北京千橡网景科技发展有限公司 | Method and equipment of sorting search results |
| CN103595747A (en) * | 2012-08-16 | 2014-02-19 | 腾讯科技（深圳）有限公司 | User-information recommending method and system |
| CN103649957A (en) * | 2011-07-20 | 2014-03-19 | 微软公司 | Modeling search in a social graph |
| CN103701836A (en) * | 2012-09-27 | 2014-04-02 | 联想(北京)有限公司 | Information processing method, terminal equipment and server |
| WO2014071782A1 (en) * | 2012-11-09 | 2014-05-15 | 腾讯科技（深圳）有限公司 | User interest recommendation method and apparatus |
| CN103853781A (en) * | 2012-12-05 | 2014-06-11 | 腾讯科技（北京）有限公司 | User group setting method and device based on social contact |
| CN104104724A (en) * | 2014-07-23 | 2014-10-15 | 深圳市富途网络科技有限公司 | Method for pushing operating mode of client side and communication system |
| CN104283723A (en) * | 2014-10-31 | 2015-01-14 | 北京蓝汛通信技术有限责任公司 | Network access log processing method and device |
| CN104281650A (en) * | 2014-09-15 | 2015-01-14 | 南京锐角信息科技有限公司 | Friend search recommendation method and friend search recommendation system based on interest analysis |
| CN104346428A (en) * | 2013-08-06 | 2015-02-11 | 索尼公司 | Information processing apparatus, information processing method, and program |
| WO2015021937A1 (en) * | 2013-08-14 | 2015-02-19 | 腾讯科技（深圳）有限公司 | Method and device for user recommendation |
| CN104468672A (en) * | 2013-09-17 | 2015-03-25 | 北京千橡网景科技发展有限公司 | Recommendation method and device for anonymous user |
| CN104462377A (en) * | 2014-12-09 | 2015-03-25 | 小米科技有限责任公司 | Contact person information providing method and device |
| CN104699859A (en) * | 2015-04-09 | 2015-06-10 | 成都卡莱博尔信息技术有限公司 | Big data collection method based on knowledge engineering |
| CN104951570A (en) * | 2015-07-27 | 2015-09-30 | 广州九尾信息科技有限公司 | Intelligent part-time job recommendation system based on data mining and LBS |
| WO2015144089A1 (en) * | 2014-03-28 | 2015-10-01 | Tencent Technology (Shenzhen) Company Limited | Application recommending method and apparatus |
| CN105027112A (en) * | 2013-03-15 | 2015-11-04 | 雅虎公司 | Page personalization based on article display time |
| CN105245378A (en) * | 2015-10-19 | 2016-01-13 | 江西掌中无限网络科技股份有限公司 | Mobile internet user relationship management method and system |
| CN105528456A (en) * | 2015-12-25 | 2016-04-27 | 北京奇虎科技有限公司 | User type based search interface showing method and device |
| CN105574033A (en) * | 2014-10-15 | 2016-05-11 | 成都理想境界科技有限公司 | Terminal, server, social network-based book sharing method and system |
| CN105635210A (en) * | 2014-10-30 | 2016-06-01 | 腾讯科技（武汉）有限公司 | Network information recommending method and device, and reading system |
| CN105741042A (en) * | 2016-02-03 | 2016-07-06 | 厦门优芽网络科技有限公司 | Automatic learning method for open approving workflow |
| CN105930399A (en) * | 2016-04-18 | 2016-09-07 | 徐亚红 | User data based dynamic contact list display method |
| WO2016165414A1 (en) * | 2015-08-28 | 2016-10-20 | 中兴通讯股份有限公司 | Method and device for push information |
| CN106055616A (en) * | 2016-05-25 | 2016-10-26 | 中山大学 | Friend recommendation method for social networking website based on named entities |
| CN103166918B (en) * | 2011-12-12 | 2016-12-14 | 阿里巴巴集团控股有限公司 | A kind of data recommendation method and device |
| CN106302748A (en) * | 2016-08-17 | 2017-01-04 | 北京字节跳动科技有限公司 | A kind of across medium personalized recommendation method and system |
| WO2017000715A1 (en) * | 2015-06-30 | 2017-01-05 | 腾讯科技（深圳）有限公司 | Contact matching method and device |
| CN106412306A (en) * | 2016-10-10 | 2017-02-15 | 珠海格力电器股份有限公司 | Social communication prompting method and device for mobile terminal and mobile terminal |
| CN106796593A (en) * | 2014-07-15 | 2017-05-31 | 微软技术许可有限责任公司 | Media are prioritized based on social data and user behavior |
| CN107395695A (en) * | 2017-07-05 | 2017-11-24 | 上海精数信息科技有限公司 | Service push method and device |
| CN107493225A (en) * | 2017-05-02 | 2017-12-19 | 朱小军 | A kind of network social intercourse method and system based on common interest |
| CN107766379A (en) * | 2016-08-22 | 2018-03-06 | 中国电信股份有限公司 | A kind of recommendation method and apparatus of Web content |
| CN107807944A (en) * | 2016-09-09 | 2018-03-16 | 北京搜狗科技发展有限公司 | A kind of file recommendation method, device and electronic equipment |
| US9984048B2 (en) | 2010-06-09 | 2018-05-29 | Alibaba Group Holding Limited | Selecting a navigation hierarchical structure diagram for website navigation |
| US10015269B2 (en) | 2014-12-09 | 2018-07-03 | Xiaomi Inc. | Method and device for providing contact information |
| CN108268660A (en) * | 2018-02-08 | 2018-07-10 | 深圳市口袋网络科技有限公司 | A kind of customer data recommends method, server and storage medium |
| CN108920531A (en) * | 2018-06-08 | 2018-11-30 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN109120719A (en) * | 2018-09-28 | 2019-01-01 | Oppo广东移动通信有限公司 | Information-pushing method, information displaying method, computer equipment and storage medium |
| CN109313638A (en) * | 2017-03-09 | 2019-02-05 | 微软技术许可有限责任公司 | Using recommendation |
| CN109918576A (en) * | 2019-01-09 | 2019-06-21 | 常熟理工学院 | A kind of microblogging concern recommended method based on joint probability matrix decomposition |
| CN110660280A (en) * | 2019-07-30 | 2020-01-07 | 湖南优积谷网络科技有限公司 | Interactive learning platform based on mobile internet |
| US10657161B2 (en) | 2012-01-19 | 2020-05-19 | Alibaba Group Holding Limited | Intelligent navigation of a category system |
| CN112732974A (en) * | 2019-10-10 | 2021-04-30 | 腾讯科技（深圳）有限公司 | Data processing method, electronic equipment and storage medium |
| CN115544386A (en) * | 2022-12-02 | 2022-12-30 | 南京新米网络科技有限公司 | Method, device and storage medium for recommending user |
| CN116107231A (en) * | 2023-04-17 | 2023-05-12 | 深圳市华曦达科技股份有限公司 | Log processing method, device and system for intelligent household equipment |
-
2006
- 2006-10-27 CN CNB200610114120XA patent/CN100530185C/en not_active Expired - Fee Related
Cited By (123)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101188521B (en) * | 2007-12-05 | 2010-07-14 | 北京金山软件有限公司 | A method for digging user behavior data and website server |
| CN101616292B (en) * | 2008-06-25 | 2012-05-23 | 索尼株式会社 | Information processing apparatus, information processing method, and information processing system |
| CN101909018A (en) * | 2009-06-05 | 2010-12-08 | 潘晓丰 | Method and system for returning to instant messaging group according to webpage browsed by user |
| CN101923545B (en) * | 2009-06-15 | 2012-10-10 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN101751448B (en) * | 2009-07-22 | 2012-06-27 | 中国科学院自动化研究所 | Commendation method of personalized resource information based on scene information |
| US9031952B2 (en) | 2009-12-31 | 2015-05-12 | Nokia Corporation | Methods and apparatuses for user interest modeling |
| CN102687166A (en) * | 2009-12-31 | 2012-09-19 | 诺基亚公司 | Methods and apparatuses for user interest modeling |
| CN102687166B (en) * | 2009-12-31 | 2016-02-10 | 诺基亚技术有限公司 | For the method and apparatus of user interest modeling |
| CN102130896A (en) * | 2010-01-14 | 2011-07-20 | 腾讯科技（深圳）有限公司 | Method and system for correlating network applications |
| CN102130896B (en) * | 2010-01-14 | 2014-05-14 | 腾讯科技（深圳）有限公司 | Method and system for correlating network applications |
| WO2011088723A1 (en) * | 2010-01-20 | 2011-07-28 | 腾讯科技（深圳）有限公司 | Method and system for recommending friends in social networking service (sns) community |
| CN102870114A (en) * | 2010-04-28 | 2013-01-09 | Jvc建伍株式会社 | Item selection device, item selection method, and item selection-use program |
| US8972419B2 (en) | 2010-04-28 | 2015-03-03 | JVC Kenwood Corporation | Item selecting apparatus, item selecting method and item selecting program |
| CN102870114B (en) * | 2010-04-28 | 2016-05-25 | Jvc建伍株式会社 | Project selecting arrangement and project selection method |
| US9740982B2 (en) | 2010-04-28 | 2017-08-22 | JVC Kenwood Corporation | Item selecting apparatus, item selecting method and item selecting program |
| US9984048B2 (en) | 2010-06-09 | 2018-05-29 | Alibaba Group Holding Limited | Selecting a navigation hierarchical structure diagram for website navigation |
| CN102279851B (en) * | 2010-06-12 | 2017-05-03 | 阿里巴巴集团控股有限公司 | Intelligent navigation method, device and system |
| CN102279851A (en) * | 2010-06-12 | 2011-12-14 | 阿里巴巴集团控股有限公司 | Intelligent navigation method, device and system |
| US9519720B2 (en) | 2010-06-12 | 2016-12-13 | Alibaba Group Holding Limited | Method, apparatus and system of intelligent navigation |
| CN101916286A (en) * | 2010-08-23 | 2010-12-15 | 宇龙计算机通信科技(深圳)有限公司 | Information recommendation method and system |
| CN101984420A (en) * | 2010-09-03 | 2011-03-09 | 百度在线网络技术（北京）有限公司 | Method and equipment for searching pictures based on word segmentation processing |
| CN101957845A (en) * | 2010-09-17 | 2011-01-26 | 百度在线网络技术（北京）有限公司 | On-line application system and implementation method thereof |
| CN102456019A (en) * | 2010-10-18 | 2012-05-16 | 腾讯科技（深圳）有限公司 | Retrieval method and device |
| CN102065454B (en) * | 2010-11-04 | 2014-07-02 | 北京邮电大学 | Context-based method for estimating amount of information |
| CN102065454A (en) * | 2010-11-04 | 2011-05-18 | 北京邮电大学 | Context-based method for estimating amount of information |
| WO2012065478A1 (en) * | 2010-11-18 | 2012-05-24 | 中兴通讯股份有限公司 | Method and apparatus for intelligent data push |
| WO2012088720A1 (en) * | 2010-12-31 | 2012-07-05 | Yahoo! Inc. | Behavioral targeted social recommendations |
| US9087106B2 (en) | 2010-12-31 | 2015-07-21 | Yahoo! Inc. | Behavior targeting social recommendations |
| WO2012122754A1 (en) * | 2011-03-15 | 2012-09-20 | 苏州阔地网络科技有限公司 | Method and apparatus for filtering dynamic information |
| CN102693223A (en) * | 2011-03-21 | 2012-09-26 | 潘燕辉 | Search method |
| WO2013004119A1 (en) * | 2011-07-05 | 2013-01-10 | 腾讯科技（深圳）有限公司 | Information acquisition method and device |
| CN102867001A (en) * | 2011-07-05 | 2013-01-09 | 腾讯科技（北京）有限公司 | Method and device for acquiring information |
| CN103649957A (en) * | 2011-07-20 | 2014-03-19 | 微软公司 | Modeling search in a social graph |
| CN102254039A (en) * | 2011-08-11 | 2011-11-23 | 武汉安问科技发展有限责任公司 | Searching engine-based network searching method |
| CN102982042B (en) * | 2011-09-07 | 2015-08-19 | 中国移动通信集团公司 | A kind of personalization content recommendation method, platform and system |
| CN102982042A (en) * | 2011-09-07 | 2013-03-20 | 中国移动通信集团公司 | Personalization content recommendation method and platform and system |
| CN103106225A (en) * | 2011-11-10 | 2013-05-15 | 北京千橡网景科技发展有限公司 | Method and device for confirming preference for user to select friends in social network |
| CN103116589A (en) * | 2011-11-17 | 2013-05-22 | 腾讯科技（深圳）有限公司 | Method and device of sending recommendation information |
| CN103136269A (en) * | 2011-12-02 | 2013-06-05 | 潘燕辉 | Method of positioning crowds according to program |
| CN103166918A (en) * | 2011-12-12 | 2013-06-19 | 阿里巴巴集团控股有限公司 | Data recommendation method and device |
| CN103166918B (en) * | 2011-12-12 | 2016-12-14 | 阿里巴巴集团控股有限公司 | A kind of data recommendation method and device |
| US10657161B2 (en) | 2012-01-19 | 2020-05-19 | Alibaba Group Holding Limited | Intelligent navigation of a category system |
| CN103218366A (en) * | 2012-01-20 | 2013-07-24 | 腾讯科技（深圳）有限公司 | Method and system for recommending download resources |
| US9264502B2 (en) | 2012-01-20 | 2016-02-16 | Tencent Technology (Shenzhen) Company Limited | Download resource recommendation method, system and storage medium |
| WO2013107238A1 (en) * | 2012-01-20 | 2013-07-25 | 腾讯科技（深圳）有限公司 | Download resource recommendation method, system and storage medium |
| CN103218366B (en) * | 2012-01-20 | 2017-09-01 | 腾讯科技（深圳）有限公司 | Download resource recommendation method and system |
| CN103327045B (en) * | 2012-03-21 | 2017-03-22 | 腾讯科技（深圳）有限公司 | User recommendation method and system in social network |
| WO2013139239A1 (en) * | 2012-03-21 | 2013-09-26 | Tencent Technology (Shenzhen) Company Limited | Method for recommending users in social network and the system thereof |
| CN103368994A (en) * | 2012-03-28 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Specific data push method and system, and retrieval server |
| CN102663073B (en) * | 2012-03-31 | 2014-08-20 | 北京奇虎科技有限公司 | Method and system for recommending based on downloaded files |
| CN102663073A (en) * | 2012-03-31 | 2012-09-12 | 奇智软件（北京）有限公司 | Method and system for recommending based on downloaded files |
| CN102685224B (en) * | 2012-04-28 | 2014-10-08 | 华为技术有限公司 | User behavior analysis method, related equipment and system |
| CN102685224A (en) * | 2012-04-28 | 2012-09-19 | 华为技术有限公司 | User behavior analysis method, related equipment and system |
| US9589275B2 (en) | 2012-04-28 | 2017-03-07 | Huawei Technologies Co., Ltd. | User behavior analysis method, and related device and method |
| WO2013159512A1 (en) * | 2012-04-28 | 2013-10-31 | 华为技术有限公司 | User behavior analysis method, and related equipment and system |
| CN103425650B (en) * | 2012-05-15 | 2018-03-16 | 腾讯科技（深圳）有限公司 | Recommend searching method and system |
| CN103425650A (en) * | 2012-05-15 | 2013-12-04 | 腾讯科技（深圳）有限公司 | Recommendation searching method and recommendation searching system |
| CN102724290A (en) * | 2012-05-23 | 2012-10-10 | 华为技术有限公司 | Method, device and system for getting target customer group |
| CN102724290B (en) * | 2012-05-23 | 2015-01-28 | 华为技术有限公司 | Method, device and system for getting target customer group |
| CN102750334A (en) * | 2012-06-01 | 2012-10-24 | 北京市农林科学院农业科技信息研究所 | Agricultural information accurate propelling method based on data mining (DM) |
| CN102750334B (en) * | 2012-06-01 | 2016-12-21 | 北京市农林科学院农业科技信息研究所 | A kind of method that Agricultural Information based on DM precisely pushes |
| CN103577486A (en) * | 2012-08-02 | 2014-02-12 | 北京千橡网景科技发展有限公司 | Method and equipment of sorting search results |
| CN103595747A (en) * | 2012-08-16 | 2014-02-19 | 腾讯科技（深圳）有限公司 | User-information recommending method and system |
| CN102855309B (en) * | 2012-08-21 | 2016-02-10 | 亿赞普(北京)科技有限公司 | A kind of information recommendation method based on user behavior association analysis and device |
| CN102855309A (en) * | 2012-08-21 | 2013-01-02 | 亿赞普(北京)科技有限公司 | Information recommendation method and device based on user behavior associated analysis |
| CN103064869A (en) * | 2012-09-05 | 2013-04-24 | 常州嘴馋了信息科技有限公司 | User profile editing system |
| CN103701836A (en) * | 2012-09-27 | 2014-04-02 | 联想(北京)有限公司 | Information processing method, terminal equipment and server |
| WO2014071782A1 (en) * | 2012-11-09 | 2014-05-15 | 腾讯科技（深圳）有限公司 | User interest recommendation method and apparatus |
| CN103810192A (en) * | 2012-11-09 | 2014-05-21 | 腾讯科技（深圳）有限公司 | User interest recommending method and device |
| CN103853781A (en) * | 2012-12-05 | 2014-06-11 | 腾讯科技（北京）有限公司 | User group setting method and device based on social contact |
| CN103853781B (en) * | 2012-12-05 | 2018-09-18 | 腾讯科技（北京）有限公司 | A kind of user's group setting method and device based on social activity |
| CN103077229A (en) * | 2013-01-04 | 2013-05-01 | 广东一一五科技有限公司 | Method and system for matching user groups |
| CN103077229B (en) * | 2013-01-04 | 2016-04-06 | 广东一一五科技有限公司 | Groups of users matching process and system |
| CN105027112B (en) * | 2013-03-15 | 2018-08-28 | 奥誓公司 | Method for page personalization |
| CN105027112A (en) * | 2013-03-15 | 2015-11-04 | 雅虎公司 | Page personalization based on article display time |
| CN104346428A (en) * | 2013-08-06 | 2015-02-11 | 索尼公司 | Information processing apparatus, information processing method, and program |
| WO2015021937A1 (en) * | 2013-08-14 | 2015-02-19 | 腾讯科技（深圳）有限公司 | Method and device for user recommendation |
| CN104468672A (en) * | 2013-09-17 | 2015-03-25 | 北京千橡网景科技发展有限公司 | Recommendation method and device for anonymous user |
| US9953262B2 (en) | 2014-03-28 | 2018-04-24 | Tencent Technology (Shenzhen) Company Limited | Application recommending method and apparatus |
| US10679132B2 (en) | 2014-03-28 | 2020-06-09 | Tencent Technology (Shenzhen) Company Limited | Application recommending method and apparatus |
| WO2015144089A1 (en) * | 2014-03-28 | 2015-10-01 | Tencent Technology (Shenzhen) Company Limited | Application recommending method and apparatus |
| CN106796593A (en) * | 2014-07-15 | 2017-05-31 | 微软技术许可有限责任公司 | Media are prioritized based on social data and user behavior |
| CN104104724A (en) * | 2014-07-23 | 2014-10-15 | 深圳市富途网络科技有限公司 | Method for pushing operating mode of client side and communication system |
| CN104104724B (en) * | 2014-07-23 | 2017-12-19 | 深圳市富途网络科技有限公司 | Method for pushing and communication system for the mode of operation of client |
| CN104281650A (en) * | 2014-09-15 | 2015-01-14 | 南京锐角信息科技有限公司 | Friend search recommendation method and friend search recommendation system based on interest analysis |
| CN105574033A (en) * | 2014-10-15 | 2016-05-11 | 成都理想境界科技有限公司 | Terminal, server, social network-based book sharing method and system |
| CN105635210A (en) * | 2014-10-30 | 2016-06-01 | 腾讯科技（武汉）有限公司 | Network information recommending method and device, and reading system |
| CN104283723A (en) * | 2014-10-31 | 2015-01-14 | 北京蓝汛通信技术有限责任公司 | Network access log processing method and device |
| CN104462377A (en) * | 2014-12-09 | 2015-03-25 | 小米科技有限责任公司 | Contact person information providing method and device |
| RU2639953C2 (en) * | 2014-12-09 | 2017-12-25 | Сяоми Инк. | Method and device for providing contact details |
| WO2016090828A1 (en) * | 2014-12-09 | 2016-06-16 | 小米科技有限责任公司 | Contact information providing method and device |
| US10015269B2 (en) | 2014-12-09 | 2018-07-03 | Xiaomi Inc. | Method and device for providing contact information |
| CN104699859A (en) * | 2015-04-09 | 2015-06-10 | 成都卡莱博尔信息技术有限公司 | Big data collection method based on knowledge engineering |
| US10841252B2 (en) | 2015-06-30 | 2020-11-17 | Tencent Technology (Shenzhen) Company Limited | Contact matching method and apparatus |
| WO2017000715A1 (en) * | 2015-06-30 | 2017-01-05 | 腾讯科技（深圳）有限公司 | Contact matching method and device |
| CN104951570A (en) * | 2015-07-27 | 2015-09-30 | 广州九尾信息科技有限公司 | Intelligent part-time job recommendation system based on data mining and LBS |
| WO2016165414A1 (en) * | 2015-08-28 | 2016-10-20 | 中兴通讯股份有限公司 | Method and device for push information |
| CN105245378A (en) * | 2015-10-19 | 2016-01-13 | 江西掌中无限网络科技股份有限公司 | Mobile internet user relationship management method and system |
| CN105528456A (en) * | 2015-12-25 | 2016-04-27 | 北京奇虎科技有限公司 | User type based search interface showing method and device |
| CN105528456B (en) * | 2015-12-25 | 2019-04-26 | 北京奇虎科技有限公司 | Search interface methods of exhibiting and device based on user type |
| CN105741042A (en) * | 2016-02-03 | 2016-07-06 | 厦门优芽网络科技有限公司 | Automatic learning method for open approving workflow |
| CN105930399A (en) * | 2016-04-18 | 2016-09-07 | 徐亚红 | User data based dynamic contact list display method |
| CN106055616A (en) * | 2016-05-25 | 2016-10-26 | 中山大学 | Friend recommendation method for social networking website based on named entities |
| CN106302748A (en) * | 2016-08-17 | 2017-01-04 | 北京字节跳动科技有限公司 | A kind of across medium personalized recommendation method and system |
| CN107766379A (en) * | 2016-08-22 | 2018-03-06 | 中国电信股份有限公司 | A kind of recommendation method and apparatus of Web content |
| CN107807944A (en) * | 2016-09-09 | 2018-03-16 | 北京搜狗科技发展有限公司 | A kind of file recommendation method, device and electronic equipment |
| CN106412306B (en) * | 2016-10-10 | 2019-11-12 | 珠海格力电器股份有限公司 | Mobile terminal social activity based reminding method, device and mobile terminal |
| CN106412306A (en) * | 2016-10-10 | 2017-02-15 | 珠海格力电器股份有限公司 | Social communication prompting method and device for mobile terminal and mobile terminal |
| CN109313638A (en) * | 2017-03-09 | 2019-02-05 | 微软技术许可有限责任公司 | Using recommendation |
| CN109313638B (en) * | 2017-03-09 | 2023-09-01 | 微软技术许可有限责任公司 | Application recommendation |
| CN107493225A (en) * | 2017-05-02 | 2017-12-19 | 朱小军 | A kind of network social intercourse method and system based on common interest |
| CN107395695A (en) * | 2017-07-05 | 2017-11-24 | 上海精数信息科技有限公司 | Service push method and device |
| CN107395695B (en) * | 2017-07-05 | 2020-09-01 | 上海精数信息科技有限公司 | Business pushing method and device |
| CN108268660A (en) * | 2018-02-08 | 2018-07-10 | 深圳市口袋网络科技有限公司 | A kind of customer data recommends method, server and storage medium |
| CN108920531A (en) * | 2018-06-08 | 2018-11-30 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN109120719A (en) * | 2018-09-28 | 2019-01-01 | Oppo广东移动通信有限公司 | Information-pushing method, information displaying method, computer equipment and storage medium |
| CN109120719B (en) * | 2018-09-28 | 2021-06-04 | Oppo广东移动通信有限公司 | Information pushing method, information display method, computer equipment and storage medium |
| CN109918576B (en) * | 2019-01-09 | 2021-01-05 | 常熟理工学院 | Microblog attention recommendation method based on joint probability matrix decomposition |
| CN109918576A (en) * | 2019-01-09 | 2019-06-21 | 常熟理工学院 | A kind of microblogging concern recommended method based on joint probability matrix decomposition |
| CN110660280A (en) * | 2019-07-30 | 2020-01-07 | 湖南优积谷网络科技有限公司 | Interactive learning platform based on mobile internet |
| CN112732974A (en) * | 2019-10-10 | 2021-04-30 | 腾讯科技（深圳）有限公司 | Data processing method, electronic equipment and storage medium |
| CN115544386A (en) * | 2022-12-02 | 2022-12-30 | 南京新米网络科技有限公司 | Method, device and storage medium for recommending user |
| CN116107231A (en) * | 2023-04-17 | 2023-05-12 | 深圳市华曦达科技股份有限公司 | Log processing method, device and system for intelligent household equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100530185C (en) | 2009-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100530185C (en) | Network behavior based personalized recommendation method and system | |
| US11281626B2 (en) | Systems and methods for management of data platforms | |
| National Research Council et al. | Frontiers in massive data analysis | |
| US10198460B2 (en) | Systems and methods for management of data platforms | |
| KR101114023B1 (en) | Content propagation for enhanced document retrieval | |
| US9626623B2 (en) | Method of automated discovery of new topics | |
| Ganti et al. | Data cleaning: A practical perspective | |
| US9390176B2 (en) | System and method for recursively traversing the internet and other sources to identify, gather, curate, adjudicate, and qualify business identity and related data | |
| US20100114899A1 (en) | Method and system for business intelligence analytics on unstructured data | |
| CN101124537A (en) | Techniques for knowledge discovery by constructing knowledge correlations using terms | |
| CN102893281A (en) | Information retrieval device, information retrieval method, computer program, and data structure | |
| CN103136360A (en) | Internet behavior markup engine and behavior markup method corresponding to same | |
| US20170109358A1 (en) | Method and system of determining enterprise content specific taxonomies and surrogate tags | |
| US11379670B1 (en) | Automatically populating responses using artificial intelligence | |
| Bellare et al. | Woo: A scalable and multi-tenant platform for continuous knowledge base synthesis | |
| CN103425740A (en) | IOT (Internet Of Things) faced material information retrieval method based on semantic clustering | |
| Pan et al. | Exploiting tractable fuzzy and crisp reasoning in ontology applications | |
| JP2004030221A (en) | Method for automatically detecting table to be modified | |
| Dong et al. | Teledata: data mining, social network analysis and statistics analysis system based on cloud computing in telecommunication industry | |
| Gözükara et al. | An Incremental Hierarchical Clustering Based System For Record Linkage In E-Commerce Domain | |
| EP3152678A1 (en) | Systems and methods for management of data platforms | |
| US10877998B2 (en) | Highly atomized segmented and interrogatable data systems (HASIDS) | |
| De Bonis et al. | Graph-based methods for Author Name Disambiguation: a survey | |
| Vakali et al. | New directions in web data management 1 | |
| Ganti et al. | Data Cleaning |
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 | ||
| ASS | Succession or assignment of patent right |
Owner name: BEIJING BONISI EDUCATION SCIENCE + TECHNOLOGY CO., Free format text: FORMER OWNER: BEIJING SOUSHEN NETWORK TECHNO Effective date: 20100705 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 100084 ROOM 4011, SANCAITANG OFFICE BUILDING, SOUTH OF TSINGHUA UNIVERSITY SOUTH GATE, HAIDIAN DISTRICT, BEIJING TO: 100081 508, 5TH FLOOR, BUILDING 3, NO.34, ZHONGGUANCUN SOUTH STREET, HAIDIAN DISTRICT, BEIJING |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20100705 Address after: 100081, No. 3, 34 South Main Street, Beijing, Zhongguancun, building 5, 508, floor, Haidian District Patentee after: Beijing Bonneth Education Technology Co., Ltd. Address before: 100084 Beijing Haidian District city on the south side of the South Gate of Tsinghua University sancaitang office room 4011 Patentee before: Beijing Soushen Network Technologies Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: ZHONG HUIBO Free format text: FORMER OWNER: BEIJING BONISI EDUCATION TECHNOLOGY CO., LTD. Effective date: 20110909 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20110909 Address after: 100081 School of humanities, Beijing Institute of Technology, 5 South Avenue, Beijing, Haidian District, Zhongguancun Patentee after: Zhong Huibo Address before: 100081, No. 3, 34 South Main Street, Beijing, Zhongguancun, building 5, 508, floor, Haidian District Patentee before: Beijing Bonneth Education Technology Co., Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090819 Termination date: 20151027 |
|
| EXPY | Termination of patent right or utility model |