CN1852124A - Client-end resource search under broadcast-storage network environment and automatic downloading method - Google Patents
Client-end resource search under broadcast-storage network environment and automatic downloading method Download PDFInfo
- Publication number
- CN1852124A CN1852124A CN 200610026690 CN200610026690A CN1852124A CN 1852124 A CN1852124 A CN 1852124A CN 200610026690 CN200610026690 CN 200610026690 CN 200610026690 A CN200610026690 A CN 200610026690A CN 1852124 A CN1852124 A CN 1852124A
- Authority
- CN
- China
- Prior art keywords
- program
- user
- client
- semantic
- classification
- 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
This invention relates to a method for carrying out resource search and automatic downloading at the customer end based on the content semantics in the broadcast-store grid environment, which puts up a bridge for negotiating web semantics between a source and sinks including that the customer end carries out three layers of semantics search matches (key information of programs, sorting match and semantics abstract information match) and automatic downloading the individual resources of users.
Description
Technical field
The invention belongs to intelligent IT technical field, be specifically related to the client-end resource search of content-based semanteme under a kind of broadcast-storage network environment and the method for downloading automatically.
Technical background
The Internet develop rapidly, the resource on the network is more and more, and present situation is " a resource surplus ", and simultaneously because magnanimity information, and the user can't obtain the information that oneself needs most, and this just need erect the bridge of communication web page semantics between the information source and the stay of two nights.And the scientific and technical means of screening content on this semantics meaning have become the new bottleneck of network media sustainable development.
Further the Internet is transformed into semantic net, allows the user search of content-based semanteme become a reality, its range of application will almost be taking the initiative in offering a hand of omnipotent contents semantic well beyond network media.Simultaneously in conjunction with user's search custom and program preferences, individual character, browsing mode realizes the automatic download of program, shows the hommization of network, has saved the time of online searching for the user.Simultaneously, under multicast environment, can efficiently solve the present situation of server end network congestion, realize " information express passway ".
Summary of the invention
The objective of the invention is to propose a kind of under broadcast-storage network environment the client-end resource search of content-based semanteme and the method for downloading automatically.
The method that the present invention proposes is used " multicast " network communication mode based on the multicast environment of " broadcast-storage network ".
Because 80/20 rule, traditional network exists the serious visit capacity and the problem of information asymmetry.80% user only visits 20% website (actual network condition is great disparity more), and promptly heavy access task is being born in the website of only a few, then " sharing conflict " problem has appearred, and serious day by day along with the sharp increase of network user's number.Adopt multicasting technology, promptly at server end, each is broadcast group and has the real-time Data Transmission ability, broadcast the content of hundreds of million every day, directly arrive the user, do not block up, it is to be allocated to need not bandwidth etc., client is selected the group of broadcasting of own required adding, and seamless connects, arbitrary program receiving.
Under this multicast environment, the present invention comprises the content of following 2 aspects:
1, client is carried out three layers of semantic retrieval coupling;
2, the user individual resource is downloaded automatically.
1. client is carried out three layers of semantic retrieval coupling
Server end cleans the program that will broadcast, and extracts the key message of program, preserve " contents semantic " of each program with the form of XML file, and client is selected own institute programs of interest and received by semantic analysis.
When the content-based semantic search of client carries out the group broadcasting program download, adopted the mode of three layers of semantic matches:
Ground floor: program key information match.
Server end cleans each program height, the concise and to the point key message that analyzes each will group broadcasting program, broadcast with programme, the user receives the group broadcasting program list, according to the key message (as keyword) in this programme content semantic information, carry out fuzzy matching and (comprise the near synonym coupling, Chinese and English coupling etc.), because user's retrieval habit and retrieval key message accuracy difference, simple utilization participle technique in the process of semantic matches, for Chinese user, it is more accurate to understand, and search efficiency is higher.
The second layer: program classification coupling
Classified information according to each program is carried out semantic matches, the mode classification of each program can be with reference to the Chinese Library abridged table of classifying, in conjunction with the catalog classification form of each website on the present network, between the accuracy of classified information and simple and easy degree, reach dynamic equilibrium, set up classification tree simultaneously.
There is the contact of interrelated interdependence in each node on the classification tree.Each node incidence coefficient between any two can be described with " incidence matrices " of one 2 dimension.Be each node and compose the ID that the overall situation is unique, represent the correlation degree of per two nodes with the matrix of one 2 dimension.Can imagine that this is a very huge matrix, but wherein most value is 0.
FTP client FTP in the enterprising lang justice of classification tree coupling, finds one or several the highest node of matching degree according to user's request.Note simultaneously, when carrying out semantic matches, the first step: carry out the match query of categorised content earlier, promptly mate or fuzzy matching, navigate to the node on the classification tree.Second step:, will have the node of certain correlation degree (can control) also to find out with the node that the first step finds by a certain threshold values by " incidence matrices ".Weight respectively in addition, by each classification tree site position with and mutual correlation degree, can obtain each program " matching degree " of this user's request.
The 3rd layer: semantic summary info coupling
Additional as to this program classification and key message, an XML file of each program also contains summary info.Summary info is carried out the method for semantic matches, be similar to full-text search, because in multicast environment, all program searchs all are to carry out in this locality, and are if all the elements of program are retrieved, obviously improper.So mode that adopts the summary info to program to mate.Set up an inverted list, summary info content process participle with program, cut speech, put into inverted list respectively, according to the result that inverted list counted, judge whether this program arrives the threshold values of user search demand (in dynamically updating), if matching degree is lower than this threshold values, then this program is not considered (in key information match, under the prerequisite that classification and matching is all failed).On summary info coupling, can be in conjunction with the algorithm of some comparative maturities, as LR, NNet (Neural network), Knn (K-nearest neighbor), SVM.
Client is by three layers of semantic retrieval coupling, the program description (XML document form) that sends to client is carried out local semantic matches, to similar close statement, Client Search Engine can be distinguished, and according to the similar degree of coupling with and correlation degree, the search of each layer is all provided its weight; For the Chinese search technology, the speech technology cut in the utilization participle, embodies the hommization retrieval service to the client.Simultaneously, to three layers of semantic matches unit, the weight difference that it is shared in Search Results, this is a dynamic parameter, and non-linear, provide three-dimensional " matching degree matrix " according to dynamic algorithm, in conjunction with three layers of semantic matches degree, from matrix, take out three layers of weight that difference is shared at every turn, obtain the search matching degree of final each program for the user, and, be shown to the user by descending; Simultaneously, a threshold values is set, the program that is lower than this matching degree value will not be shown to the user, and this threshold values can be disposed by the user.System also can download the statistics of situation according to the program of user's Search Results in the past, obtains a threshold values through algorithm, is the reference of user institute.Restriction program range of choice provide to meet the high program of the user preferences degree of coupling, thereby embodying the user serves hommization on the one hand, saves user's browsing time, on the other hand, saves the client expense, reaches most effective.
Be M=K*k%+C*c%+A*a% (＞T)
Here M representative: for a certain particular demands of user, the whole matching degree (match) of a certain file.K represents the keyword matching degree (keyword match) of this document, and on behalf of keyword, k% mate proportion shared in the whole matching degree; The matching degree (category match) of C representative classification, c% represents classification and matching shared proportion in the whole matching degree; A represents the matching degree (abstract match) of summary info, and a% represents digests match shared proportion in the whole matching degree.Here T is set threshold values, just directly the abandoning of those matching degrees very low (such as＜60%), and the file that comes out as a matched and searched is selected for the user.Client is carried out semantic matches, and classification tree module and Dynamic matching matrix module are transmitted by the specific group of broadcasting between service centre and the client, guarantee synchronously.
This method has been taked the network transmission format of multicast, and all search work are finished at client terminal local.
2. the user individual resource is downloaded automatically
Specific user's hobby always has certain continuity, has new program to occur at every turn, and the user need re-enter user's search information originally, carries out the selection of program, and this is repetitive operation to the user.Client, individual subscriber hobby storage form is provided on the one hand, on the other hand, system watches the custom of program for a long time by recording user, sum up the preference degree of user's various types of programs, the program that positive representation user is high to preference degree (program of preferring) is implemented the predetermined of program and is downloaded then.Make the user needn't go again to seek the program of oneself liking, but the program that allows the user like is looked for the user automatically.
The user is as follows to definite mode of program preference degree: the search information of the each input of system log (SYSLOG) user, the each formative program search result (XML file format) who draws according to semantic matches, cooperate the matching degree threshold values that the user is set or system-computed goes out simultaneously, it is interested in the program of the program of which classification, which key message to draw the user.System simultaneously recording user to the operation of program downloading, browse frequency, retention time etc., if be user's hobby place really, then this type of program will be enjoyed higher " user preference ".If after the user's download immediately the deletion or access frequency is very low, then the user is not high to the actual interest of this program.In user's search matching degree and program is actual watches on the rate, be one of each classification program distribution " user preferences degree " after system's weighting promptly, this value is definite according to the testing result of user behavior by system, constantly is among the dynamic process.
Simultaneously, server end is given in the client passback that the statistical information of user behavior is regular.Server receives the hobby from each user, habit statistical, at server end, analyze according to statistical model, the structure of suitable adjustment " classification tree ", if some classification most of users all lose interest in, lower then that this is sorted in position in the classification tree, make it more to approach a page node; To all very interested content of those most of users, suitably improve its position in classification tree, make it to move to root node.In addition, for those practical works, programs such as silver screen hits can suitably be adjusted its keyword or classification, and for example " focus in the recent period ", with the pouplarity of outstanding this program, these programs upgrade in time owing to ageing needs simultaneously; Equally, some unfailing classics can solely be warded off a classification or titled with keyword " classics " " treasured book " and so on.These feedbacks are broadcasted also for the program of server end foundation are provided, thereby it is more effective that program is broadcasted, and reach better " audience ratings ".
Like this, when the user receives programme, can seek current the most popular and be everybody confessed program with a definite target in view.Effectively help the user in the information of magnanimity, sought the program of most worthy, obtained maximum amount of information.
Description of drawings
Fig. 1 is the structure of original classification tree.
Fig. 2 is the structure through the classification tree after the user preferences feedback.
Embodiment
Further specify the actual mechanical process of said method in conjunction with a concrete client implementation example.
Under the broadcast-storage network system, client is at first downloaded the programme on the specific channel and the necessary files such as semantic classification trees of service centre.
1, the user at first checks programme (classification of programme this moment is according to the structure of original classification tree), selects own programs of interest.This moment, the user can directly select the program oneself liked, and perhaps the interested content of input is mated by three layers of semantic retrieval.
2, the user at first imports own interested resource information, such as " football ", and " computer studies ", " the 10th National Games ", " hot topic " etc.At this moment, FTP client FTP carries out " three layers of semantic retrieval coupling " to all resources in the resource programme.At first carry out the coupling retrieval of key message, draw the resource program cells of its matching degree on a certain threshold values.Secondly, carry out the classification and matching retrieval.The interested content that fuzzy match user is imported on classification tree.Such as, on classification tree, finding the branch node of " football ", those resources that belong to " football " of classifying in so all resource programmes are that user institute is interested, and those are picked out, row are given the user candidate.Simultaneously, on the classification tree, " football " node degree of coupling is than those higher nodes, such as " English soccer team ", the pairing resource of " Mr. football " classification also probably is that the user relatively is concerned about, those resources also should be given low slightly matching degree, offer the user as the search matching result and select.System also can carry out the coupling of semantic summary info, finds out corresponding resource.Matching degrees of these search return results are given weighting, can obtain all promptly will multicast (displaying by screens of some) resource for the matching degree of user preference.After the ordering, the user can see that system offers user's satisfactory resource program, and the user selects again.
Such as, the user has imported " race ", and " track and field ", classify and summary info with reference to the keyword of each program in the programme this moment, and consider the structure of semantic classification trees.According to M=K*k%+C*c%+A*a% (＞can find that T) algorithm is as follows:
In the programme, find file F1, it has a keyword to be " track and field "; F2, it has a keyword to be " track and field "; F3, it has a keyword to be " track and field " respectively, and a keyword is " race " in addition; Assignment K1=50% then, K2=50%, K3=90%.
Then check classified information, can see by the classification chart of Fig. 1 that those files under track and field contest matter are to meet user's request most, that finds F1 is categorized as " track and field ", then assignment C1=60%; F2 is categorized as " track star " then assignment C2=50%; F3 is categorized as " track and field contest matter ", assignment C3=90%; Find to have be categorized as " physical culture " of a file F4 simultaneously, it is certain related to see that sport category and track and field class have this moment, check " incidence matrices ", the relating value of finding track and field and physical culture is 70%, then at this moment, assignment C4=60%*70%=42%, (having adopted the computational methods of the classification associated degree of simple classification and matching degree * this moment).
The coupling work of making a summary is at last carried out the LR digests match with the clip Text of each file, draws to file F1 its A1=60%; To file F2, its A2=50%, to file F3, its A3=80%, to file F4, its A4=40%; Find to have file F5, its A5=45% again.The digests match of remaining paper is not considered all less than 40%.
Add up at last, to native system, the distribution between the weights: k%=40%, c%=35%, a%=25% is 40% according to the set threshold values of this this machine of user system simultaneously, promptly matching degree＜60% is not shown to the user, directly abandons.
According to: M=K*k%+C*c%+A*a% (＞T)
Calculate M1=50%*40%+60%*35%+60%*25%=56%＞40% (selecting) for the user
M2=50%*40%+50%*35%+50%*25%=50%＞40% (selecting) for the user
M3=90%*40%+90%*35%+80%*25%=87.5%＞40% (selecting) for the user
M4=0.00*40%+42%*35%+40%*25%=24.7%＜40% (not being the candidate)
M5=0.00*40%+0.00*35%+45%*25%=11.25%＜40% (not being the candidate)
Because the semantic matches of this moment is finished in client, semantic tree etc. need client and service central synchronous, need be in the downloaded resources programme, and more newly downloaded synchronously on specific channel.
3, simultaneously, the automatic recording user of system is each oneself selects to want the key message of program downloading, as the keyword of this program, classification etc., and the interested resource information of each user's input, and each return results of system and user's secondary selects, and these all are the key messages of user preference.If the user had the information of search " football " in continuous several days, we determine that relatively this user is a football fan so, if he downloads the information of Real Madrid team at every turn, we think that he is likely the football fan of Real Madrid so.So, after several days statistics of system, during system's " user preferences " preserves, will be automatically for the user add football, the key message of Real Madrid, if the football that then each system comes out for this user search, the resource of Real Madrid, this user has the download of selection, and the user preferences degree that the football Real Madrid then is described is than higher,, be 90 then such as assignment with dynamic user preferences degree.Spent several days again, found that the user is not selected to download to the resource program of football, Real Madrid, then preference degree descends dynamically, such as dropping to 60.If several days again, the user was still not interested in this, preference degree continues to descend, and drops to certain threshold values, and such as 45, then this type of content is no longer searched for automatically for the user by system.Simultaneously the user also can the interested content of own predefined oneself, allows the system all be that the user searches for automatically at every turn.
4, because server end is given in the client passback that the statistical information of user behavior is regular, server end is according to adding up the recent incident personage that everybody relatively is concerned about etc., the structure of suitable adjustment " classification tree ", such as, originally " Liu Xiang " was placed on " physical culture "-" track and field "-" track star "-" Liu Xiang ", server end is in time adjusted now, " Liu Xiang " is placed in " focus personage " classification, then client search " focus in the recent period ", during key messages such as " person who attract people's attentions ", by fuzzy matching, can find out the information of current " Liu Xiang " that everybody is concerned about.(referring to Fig. 1, Fig. 2)
Claims (3)
1, a kind of client-end resource search under broadcast-storage network environment and automatic downloading method is characterized in that the multicast environment based on broadcast-storage network, utilization multicast network communication pattern, and concrete steps are:
(1) client is carried out three layers of semantic retrieval coupling:
Server end cleans the program that will broadcast, and extracts the key message of program, preserve the contents semantic of each program with the form of XML file, and client is selected own institute programs of interest to receive by semantic analysis;
When the content-based semantic search of client carries out the group broadcasting program download, adopted the mode of three layers of semantic matches:
Ground floor: program key information match
Server end cleans each program, and the concise and to the point key message that analyzes each will group broadcasting program broadcasts with programme, and the user receives the group broadcasting program list, and the key message according in this programme content semantic information carries out fuzzy matching;
The second layer: program classification coupling
Classified information according to each program is carried out semantic matches, sets up classification tree simultaneously, and each node incidence coefficient is between any two described with the incidence matrices of one 2 dimension on the classification tree;
FTP client FTP is according to user's request, in the enterprising lang justice of classification tree coupling, the first step: carry out the match query of categorised content earlier, promptly mate or fuzzy matching, navigate to the node on the classification tree; Second step: pass through incidence matrices, to there be the node of certain correlation degree also to find out with the node that the first step finds, weight respectively in addition, by each classification tree site position with and mutual correlation degree, obtain each program " matching degree " of this user's request;
The 3rd layer: semantic summary info coupling
Set up an inverted list,, cut speech the summary info content process participle of program, put into inverted list respectively,, judge whether this program arrives the threshold values of user search demand according to the result that inverted list counted, if matching degree is lower than this threshold values, then this program is not considered;
(2) download automatically with personalized resource:
Client, individual subscriber hobby storage form is provided on the one hand, and on the other hand, system watches the custom of program for a long time by recording user, sum up the preference degree of user's various types of programs, positive representation user implements the predetermined of program to the high program of preference degree and downloads then.
2, method according to claim 1, it is characterized in that described user is as follows to definite mode of program preference degree: the search information of the each input of system log (SYSLOG) user, the each formative program search result who draws according to semantic matches, cooperate the matching degree threshold values that the user is set or system-computed goes out simultaneously, it is interested in the program of the program of which classification, which key message to draw the user; System simultaneously recording user to the operation of program downloading, browse frequency, retention time etc., if be user's hobby place really, then this type of program will be enjoyed higher " user preference "; If after the user's download immediately the deletion or access frequency is very low, then the user is not high to the actual interest of this program.
3, method according to claim 2 is characterized in that the client passback that the statistical information of user behavior is regular is to server end; Server receives the hobby from each user, and habit statistical is at server end, analyze according to statistical model, the structure of suitable adjustment classification tree is if the most of users of some classification lose interest in, then loweing, this is sorted in position in the classification tree, makes it more to approach a page node; To all very interested content of those most of users, suitably improve its position in classification tree, make it to move to root node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100266903A CN100384134C (en) | 2006-05-18 | 2006-05-18 | Client-end resource search under broadcast-storage network environment and automatic downloading method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100266903A CN100384134C (en) | 2006-05-18 | 2006-05-18 | Client-end resource search under broadcast-storage network environment and automatic downloading method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1852124A true CN1852124A (en) | 2006-10-25 |
| CN100384134C CN100384134C (en) | 2008-04-23 |
Family
ID=37133595
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100266903A Expired - Fee Related CN100384134C (en) | 2006-05-18 | 2006-05-18 | Client-end resource search under broadcast-storage network environment and automatic downloading method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100384134C (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101170578B (en) * | 2007-11-30 | 2010-06-30 | 北京理工大学 | Hierarchical peer-to-peer network structure and constructing method based on syntax similarity |
| CN102006320A (en) * | 2010-11-08 | 2011-04-06 | 北京大学 | Grid geographic information service matching method based on semantics |
| WO2011079443A1 (en) * | 2009-12-30 | 2011-07-07 | Nokia Corporation | Intelligent reception of broadcasted information items |
| CN101442560B (en) * | 2008-12-10 | 2011-08-03 | 南京邮电大学 | Method for playing and enquiring mobile communication terminal image and sound based on mobile grid |
| CN102156695A (en) * | 2010-02-11 | 2011-08-17 | 泰怡凯电器(苏州)有限公司 | Automatic information collection device and working method thereof |
| WO2012094965A1 (en) * | 2011-01-12 | 2012-07-19 | 联想(北京)有限公司 | Method, terminal and server for presenting prompt message |
| CN102012900B (en) * | 2009-09-04 | 2013-01-30 | 阿里巴巴集团控股有限公司 | An information retrieval method and system |
| CN101606174B (en) * | 2007-02-05 | 2013-11-06 | 微软公司 | Techniques to manage a taxonomy system for heterogeneous resource domains |
| CN106489141A (en) * | 2014-08-21 | 2017-03-08 | 三星电子株式会社 | For method that content is classified and electronic equipment |
| CN107133271A (en) * | 2017-04-01 | 2017-09-05 | 上海半坡网络技术有限公司 | The semantic real-time expression system of mind map and its operating method |
| CN112650878A (en) * | 2019-10-11 | 2021-04-13 | 北京声智科技有限公司 | Retrieval method, system, device and medium |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103581255A (en) * | 2012-08-02 | 2014-02-12 | 百度在线网络技术（北京）有限公司 | Content source analyzing method, system and device |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7693827B2 (en) * | 2003-09-30 | 2010-04-06 | Google Inc. | Personalization of placed content ordering in search results |
| JP4124115B2 (en) * | 2003-12-02 | 2008-07-23 | ソニー株式会社 | Information processing apparatus, information processing method, and computer program |
| CN1604643A (en) * | 2004-11-04 | 2005-04-06 | 复旦大学 | A method for automatic downloading video program according to users' taste |
| US7529733B2 (en) * | 2004-11-10 | 2009-05-05 | International Business Machines Corporation | Query builder using context sensitive grids |
-
2006
- 2006-05-18 CN CNB2006100266903A patent/CN100384134C/en not_active Expired - Fee Related
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101606174B (en) * | 2007-02-05 | 2013-11-06 | 微软公司 | Techniques to manage a taxonomy system for heterogeneous resource domains |
| CN101170578B (en) * | 2007-11-30 | 2010-06-30 | 北京理工大学 | Hierarchical peer-to-peer network structure and constructing method based on syntax similarity |
| CN101442560B (en) * | 2008-12-10 | 2011-08-03 | 南京邮电大学 | Method for playing and enquiring mobile communication terminal image and sound based on mobile grid |
| CN102012900B (en) * | 2009-09-04 | 2013-01-30 | 阿里巴巴集团控股有限公司 | An information retrieval method and system |
| WO2011079443A1 (en) * | 2009-12-30 | 2011-07-07 | Nokia Corporation | Intelligent reception of broadcasted information items |
| US20120297011A1 (en) * | 2009-12-30 | 2012-11-22 | Nokia Corporation | Intelligent Reception of Broadcasted Information Items |
| CN102156695A (en) * | 2010-02-11 | 2011-08-17 | 泰怡凯电器(苏州)有限公司 | Automatic information collection device and working method thereof |
| CN102006320B (en) * | 2010-11-08 | 2013-04-17 | 北京大学 | Grid geographic information service matching method based on semantics |
| CN102006320A (en) * | 2010-11-08 | 2011-04-06 | 北京大学 | Grid geographic information service matching method based on semantics |
| WO2012094965A1 (en) * | 2011-01-12 | 2012-07-19 | 联想(北京)有限公司 | Method, terminal and server for presenting prompt message |
| CN106489141A (en) * | 2014-08-21 | 2017-03-08 | 三星电子株式会社 | For method that content is classified and electronic equipment |
| CN107133271A (en) * | 2017-04-01 | 2017-09-05 | 上海半坡网络技术有限公司 | The semantic real-time expression system of mind map and its operating method |
| CN107133271B (en) * | 2017-04-01 | 2021-04-06 | 上海半坡网络技术有限公司 | Semantic brain graph real-time expression system and operation method thereof |
| CN112650878A (en) * | 2019-10-11 | 2021-04-13 | 北京声智科技有限公司 | Retrieval method, system, device and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100384134C (en) | 2008-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100384134C (en) | Client-end resource search under broadcast-storage network environment and automatic downloading method | |
| US8176029B2 (en) | Composite display method and system for search engine of same resource information based on degree of attention | |
| Airoldi et al. | Follow the algorithm: An exploratory investigation of music on YouTube | |
| CN103488788B (en) | It is a kind of to apply the method and device recommended automatically | |
| CN103023971A (en) | Information pushing method and system of music sharing radio stations | |
| TWI475412B (en) | Digital content reordering method and digital content aggregator | |
| CN103428267B (en) | A kind of wisdom caching system and the method distinguishing user preferences dependency thereof | |
| CN103744849B (en) | Method and device for automatic recommendation application | |
| CN102779308A (en) | Advertisement release method and system | |
| CN103634687B (en) | The method and system of video search result are provided in intelligent television | |
| US8527450B2 (en) | Apparatus and methods for analyzing and using short messages from commercial accounts | |
| CN102073699A (en) | Method, device and equipment for improving search result based on user behaviors | |
| Nasraoui et al. | A framework for mining evolving trends in web data streams using dynamic learning and retrospective validation | |
| CN101232415A (en) | Equity network node visit apparatus, method and system | |
| CN1912875A (en) | Method of file distributing and searching and its system | |
| Luu et al. | Alvis peers: a scalable full-text peer-to-peer retrieval engine | |
| CN113656673A (en) | Master-slave distributed content crawling robot for advertisement delivery | |
| CN102214227A (en) | Automatic public opinion monitoring method based on internet hierarchical structure storage | |
| CN102331929A (en) | Service classification and recommendation method based on service combination history | |
| CN105354343B (en) | User characteristics method for digging based on remote dialogue | |
| CN102298621B (en) | System for obtaining page user focus degree PageFocus by method for aggregating and displaying same source information search engine based on focus degree | |
| Shavitt et al. | Song clustering using peer-to-peer co-occurrences | |
| CN1547351A (en) | Collaborative filtering recommendation approach for dealing with ultra-mass users | |
| Yadav et al. | Parallel crawler architecture and web page change detection | |
| Kitamoto | Digital typhoon: Near real-time aggregation, recombination and delivery of typhoon-related information |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080423 Termination date: 20110518 |