CN102306182A - Method for excavating user interest based on conceptual semantic background image - Google Patents
Method for excavating user interest based on conceptual semantic background image Download PDFInfo
- Publication number
- CN102306182A CN102306182A CN201110252770A CN201110252770A CN102306182A CN 102306182 A CN102306182 A CN 102306182A CN 201110252770 A CN201110252770 A CN 201110252770A CN 201110252770 A CN201110252770 A CN 201110252770A CN 102306182 A CN102306182 A CN 102306182A
- Authority
- CN
- China
- Prior art keywords
- user
- notion
- system server
- concept
- interest
- 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.)
- Pending
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The invention relates to the field of network research, and discloses a method for excavating user interest based on a conceptual semantic background image. The method comprises the following steps that: (A) a system server acquires user theme data; (B) the system server establishes a user theme interest model; (C) the system server establishes the conceptual semantic background image; and (D) the system server updates the conceptual semantic background image to update user theme interest data. The method has the advantages that: the conceptual semantic background image is provided, so that a network crawler can better select a crawling direction, and the method can be used for more quickly and efficiently finding a web page interested by a user in vast Internet ocean compared with a breadth-first theme crawling method and the conventional theme crawling method; meanwhile, similarity calculation is brought forward to a conceptual level, and semantic matching is performed, so that the relevance of the user interest and the web page can be better calculated.
Description
Technical field
The present invention relates to the web search field, relate in particular to a kind of method based on the semantic Background digging user of notion interest.
Background technology
There are two kinds with the relevant theme of the semantic Background of the notion aspect of creeping: the one, link Background; M.Diligenti; F.M.Coetzee; S.Lawrence; C.L.Giles; Focusedcrawlingusingcontextgraphs; The26thInternationalConferenceonVeryLargeDatabase(VLDB); 2000; Pp.527 – 534. the method are with among webpage relationship map to a figure on the network; Form the link Background of webpage, can be used for judging the distance between webpage to be creeped and user's the target web, and then arrange creeping in proper order of webpage.But the method for link Background is based on a kind of hypothesis, between all webpages in same theme a kind of hierarchical relationship is arranged all, yet when not having this hierarchical relationship between the webpage, the method for link Background can't be used.The 2nd, background context figure; H.Ching-Chi; W.Fan; Topic-specific crawling on the Web with the measurements of the relevancy context graph; Information Systems 31 (2006) 232 – 246. this methods are a kind of improvement to link Background method; It is based on and proposes under this hypothesis; Be linked to those webpages of same webpage; They often have relevant content; Semantic some relevant webpages, they also can be linked on some relevant webpages of content.But these two kinds of methods just rest on the aspect of keyword matching the judgement of similarity.And the semantic Background of notion has advanceed to calculation of similarity degree on the aspect of notion, makes coupling semantically, can calculate the degree of correlation of user interest and webpage better.
Summary of the invention
In order to solve the problems of the prior art, the invention provides a kind of method based on the semantic Background digging user of notion interest, solve in the prior art that the judgement of similarity only rests on the problem on the keyword matching aspect in network principal is creeped.
The invention provides a kind of method based on the semantic Background digging user of notion interest, may further comprise the steps: (A) system server is gathered user's subject data; (B) system server is set up user's theme interest model; (C) system server is set up the semantic Background of notion; (D) system server the more semantic Background of new ideas to upgrade user's theme interesting data.
As further improvement of the present invention, in the said step (A), system server is submitted to search engine to user's query word and in the result who returns page or leaf, is allowed the user select the webpage of being interested in and constitute user's theme interest set.
As further improvement of the present invention, in the said step (B), system server is built the concept map of upright reflection user interest jointly through the theme interest set; Said system server made up concept lattice earlier before setting up user's theme interest model.
As further improvement of the present invention, in the said step (C), said system server converts concept lattice into the semantic Background of notion of the semantic relation between can the visual representation webpage.
As further improvement of the present invention, in the said step (D), said system server increases or reduces the semantic Background of notion.
As further improvement of the present invention; Said system server with the concrete grammar that concept lattice converts the semantic Background of notion of the semantic relation between can the visual representation webpage into is: system server is confirmed key concept earlier and key concept is put into the 0th layer of conceptual backdrop figure; And then utilize what of attribute number of each notion in the concept lattice, the notion in the concept lattice is mapped in the middle of the corresponding level of conceptual backdrop figure.
As further improvement of the present invention, system server utilizes the attribute speech of concept lattice to give the notion layering in the concept lattice, and its layered approach is: the key concept in the concept lattice is inserted among the conceptual backdrop figure, as the 0th layer of conceptual backdrop figure; In the non-core notion of concept lattice, comprise the attribute speech of key concept fully and the node of the attribute speech of higher similarity as the ground floor of conceptual backdrop figure arranged with key concept; In the remaining non-core concept set, comprise the notion of the attribute speech of (N-i+1) individual key concept, as the i layer of conceptual backdrop figure, wherein N representes the number of the attribute speech in the key concept, i ∈ [1, N].
As further improvement of the present invention; System server is collected the webpage that the user selects; Form the interest topic collections of web pages; Again collections of web pages is carried out word segmentation processing, feature extraction; And calculated characteristics weights; Then these characteristics are sorted according to the weights size, the characteristic of selecting to come the front is as the community set that makes up concept lattice.
As further improvement of the present invention, said community set comprises artificial intelligence, machine learning, Knowledge Discovery, agent technology, pattern-recognition, natural language processing, feature extraction, mechanical translation and the representation of knowledge.
The invention has the beneficial effects as follows: the proposition of the semantic Background of notion; Help the network crawl worm and select the direction of creeping better; Compare with traditional theme method of creeping with breadth First, it can find the user's interest webpage sooner, more efficiently in the ocean, internet of vastness; Simultaneously, it has advanceed to calculation of similarity degree on the aspect of notion, carries out coupling semantically, can calculate the degree of correlation of user interest and webpage better.
[description of drawings]
Fig. 1 is the process flow diagram that the present invention is based on the method for the semantic Background digging user of notion interest.
Fig. 2 is an enforcement illustration of concept lattice among the present invention.
Fig. 3 is that the semantic Background of user's theme interest among the present invention is implemented illustration.
Fig. 4 is the process flow diagram that concept lattice converts conceptual backdrop figure among the present invention.
Fig. 5 is the process flow diagram that increases the notion among the conceptual backdrop figure among the present invention.
Fig. 6 is the process flow diagram that reduces the notion among the conceptual backdrop figure among the present invention.
[embodiment]
Below in conjunction with description of drawings and embodiment the present invention is further specified.
As shown in Figure 1, a kind of method based on the semantic Background digging user of notion interest, may further comprise the steps: (A) system server is gathered user's subject data; (B) system server is set up user's theme interest model; (C) system server is set up the semantic Background of notion; (D) system server the more semantic Background of new ideas to upgrade user's theme interesting data.
In the said step (A), system server is submitted to search engine to user's query word and in the result who returns page or leaf, is allowed the user select the webpage of being interested in and constitute user's theme interest set.
In the said step (B), system server is built the concept map of upright reflection user interest jointly through the theme interest set; Said system server made up concept lattice earlier before setting up user's theme interest model.
In the said step (C), said system server converts concept lattice into the semantic Background of notion of the semantic relation between can the visual representation webpage.
In the said step (D), said system server increases or reduces the semantic Background of notion.
Said system server with the concrete grammar that concept lattice converts the semantic Background of notion of the semantic relation between can the visual representation webpage into is: system server is confirmed key concept earlier and key concept is put into the 0th layer of conceptual backdrop figure; And then utilize what of attribute number of each notion in the concept lattice, the notion in the concept lattice is mapped in the middle of the corresponding level of conceptual backdrop figure.
System server utilizes the attribute speech of concept lattice to give the notion layering in the concept lattice, and its layered approach is: the key concept in the concept lattice is inserted among the conceptual backdrop figure, as the 0th layer of conceptual backdrop figure; In the non-core notion of concept lattice, comprise the attribute speech of key concept fully and the node of the attribute speech of higher similarity as the ground floor of conceptual backdrop figure arranged with key concept; In the remaining non-core concept set, comprise the notion of the attribute speech of (N-i+1) individual key concept, as the i layer of conceptual backdrop figure, wherein N representes the number of the attribute speech in the key concept, i ∈ [1, N].
System server is collected the webpage that the user selects; Form the interest topic collections of web pages, again collections of web pages is carried out word segmentation processing, feature extraction, and the calculated characteristics weights; Then these characteristics are sorted according to the weights size, the characteristic of selecting to come the front is as the community set that makes up concept lattice.
Said community set comprises artificial intelligence, machine learning, Knowledge Discovery, agent technology, pattern-recognition, natural language processing, feature extraction, mechanical translation and the representation of knowledge.
User's subject data is gathered:
Use the semantic Background of notion, at first, make up user's theme interest, the mode that adopts the user to participate in is usually confirmed user's interest.As user's query word is submitted to search engine GOOGLE; In ten results of first page that it returns, allow the user select the webpage of being interested in; Constitute user's theme interest set, remove to set up user's topic model, promptly reflect the concept map of user interest through this collections of web pages.
Set up user's theme interest model:
Obtained after the set of user interest subject web page, need utilize these pages to come to set up this theme of theme feature relational model performance for the user.What the present invention adopted is the method for form concept analysis, shows user's ferret out background through making up concept lattice.At first that the user is selected group of web lumps together; Form the interest topic collections of web pages; Again collections of web pages is carried out word segmentation processing, feature extraction; And calculated characteristics weights; Again these characteristics are sorted according to the weights size; Selection comes N characteristic of front as the community set that makes up concept lattice, sets up model then, and the process of modeling at first is to make up concept lattice.For example, the user has selected the characteristic shown in a ~ i as property set, and 8 pages have constituted the form context table that is shown in the following figure as object set, and then makes up the concept lattice shown in figure two.
A: artificial intelligence b: machine learning
C: Knowledge Discovery d:agent technology
E: pattern-recognition f: natural language processing
G: feature extraction h: mechanical translation
I: the representation of knowledge
Object?a b c d e f g h i
1 * * ?*
2 * * ?* *
3 * * * * *
4 * * * * *
5 * * * *
6 * * * * *
7 * * * *
8 * * * *
Set up the semantic Background of notion:
After having set up the concept lattice of user's interest, it is converted into the semantic Background of notion of the semantic relation between can the visual representation webpage.Method for transformation is following: at first will determine key concept, the theme feature collection of supposing user's appointment is " abc ", so just can be that the notion of " abc " regarded key concept as with property set, key concept is put into conceptual backdrop figure the 0th layer.And then utilize the attribute of each notion in the concept lattice to comprise what of key concept attribute number; Notion in the lattice is mapped in the middle of the corresponding level of Background; Principle is following; If the number of the attribute speech of key concept is N; In the then non-core concept set; The attribute speech comprises the number of the speech of key concept and puts into the i layer for those individual attributes of (N-i+1), i ∈ [1, N].All notions in concept lattice all are mapped in the Background, have just constituted the semantic Background model that can represent the user interest theme.Semantic Background such as Fig. 3, shown in Figure 4.
Conceptual backdrop figure uses:
After semantic Background is set up; Just can judge the similarity of wait to creep webpage and theme with it; In the time of a new web page; This web page contents is carried out word segmentation processing, and calculate the weighted value of the attribute speech that extracts, and represent this webpage with these attribute speech with the TF-IDF method; Do the calculating of semantic similarity again with the notion of conceptual backdrop figure; The formula that calculates concept similarity is following: defined notion (E1, I1) with (E2, I2) similarity calculating method between is following:
SimCC((E1,I1),(?E2,I2))=
|(?E1∩E2)|*w+
|?(I1∩I2)|*(1-w)
r m
Wherein r is the number maximal value of element among set E1 and the E2, m be set I1 with I2 in the maximal value of number of element, w is a weight, E1 ∩ E2 is illustrated in and gathers identical element number among E1 and the E2, I1 ∩ I2 is illustrated in the number of gathering element identical among I1 and the I2.
In the process of creeping, need dynamically update conceptual backdrop figure, with better expression user's interest.Use the theme of the method for conceptual backdrop figure to creep, can improve recall rate and the accurate rate of theme reptant effectively, improve the efficient of creeping.
Be elaborated in the method process flow diagram below of renewal conceptual backdrop figure.
In one embodiment, for the service of a satisfaction can be provided to the user, just must accurately learn user's interest place.Generally can carry out the search of key word, the theme interest of in the result set that returns, going to follow the tracks of the user through universal search engine.The present invention returns the webpage relevant with the keyword of user's initial input through the Web Service interface that Google provides; And with these results that return as candidate's subject web page; Recommend the user; The user can do mark down to its satisfied webpage in navigation process, these pages have just constituted user's theme interest set.
After having obtained the set of user interest subject web page, just need utilize these webpages to come for the user sets up a theme feature relational model, this model is wanted well to show this theme.What this paper used is the method for form concept analysis, through setting up the ferret out background that concept lattice shows the user.At first that the user is selected group of web lumps together; Form the interest topic collections of web pages; These webpages are carried out word segmentation processing extract the collections of web pages characteristic; And adopt the TF-IDF method that characteristic is carried out the weighting statistics; These characteristics are sorted according to the weights size, N the characteristic of selecting to come the front is as the community set that makes up concept lattice again.
Calculate the similarity between webpage to be creeped and the user's theme interest with conceptual backdrop figure.Experiment showed, that the attribute speech that utilizes notion is obvious to the notion layered effect in the concept lattice, layered approach is following:
Therefore 1) key concept can reflect user's theme interest, the key concept in the concept lattice is inserted into the centre of conceptual backdrop figure, as the 0th layer of conceptual backdrop figure.
2) in the non-core notion in concept lattice, the attribute speech of notion comprises the attribute speech of key concept fully, with key concept higher similarity is arranged, as the node in the ground floor of conceptual backdrop figure.
3) in remaining non-core concept set, the attribute speech of notion comprises those notions of the attribute speech of (N-i+1) individual key concept, and as the I layer of conceptual backdrop figure, wherein N representes the number of the attribute speech in the key concept, i ∈ [1, N].
Each notion in the concept lattice is mapped in the corresponding level of conceptual backdrop figure and goes, form conceptual backdrop figure.
Increment concept updating conceptual backdrop figure: like Fig. 5,
1) the increment notion generates step:
1, at first according to the level N of these new ideas of attribute number judgment in Background of new object;
2, then all concept attributes on the N-1 layer in the attribute of new object and the Background are sought common ground;
3,, draw dissimilar increment notions according to common factor result's difference;
2) incrementally updating conceptual backdrop figure
A. find out the relevant webpage of theme
The prediction score big webpage relevant webpage that promptly is the theme
B. upgrade conceptual backdrop figure
Step of updating is following:
(i) at first judge the level N of increment notion.
(ii) calculate the similarity of all notions on this increment notion and the N-1 layer then, get the sub-notion (internal layer be sub-notion, skin be father notion) of that maximum notion of similarity, simultaneously newly-increased limit between these two notions as this increment notion.
(iii) all join among the conceptual backdrop figure up to all increment notions.Arthmetic statement is following:
Reduce concept updating conceptual backdrop figure: like Fig. 6,
Along with the time constantly changes; The theme reptant climbs in the webpage of returning and has some out-of-date webpages; These webpages or content can not perhaps not exist by fine reflection user theme; Can there be some information out-of-date, that can not reflect user's theme in the corresponding concept Background so; At this moment just need in conceptual backdrop figure, in time delete these out-of-date information; These outdated informations are with the embodied of notion in conceptual backdrop figure, reach the purpose of deletion outdated information through some notions among the deletion conceptual backdrop figure.Step is following:
(1) from the low webpage of prediction score, finds and the incoherent webpage of theme.
(2) the deletion notion relevant with uncorrelated webpage.
Use the theme of the method for conceptual backdrop figure to creep, can improve recall rate and the accurate rate of theme reptant effectively, improve the efficient of creeping.
Above content is to combine concrete preferred implementation to the further explain that the present invention did, and can not assert that practical implementation of the present invention is confined to these explanations.For the those of ordinary skill of technical field under the present invention, under the prerequisite that does not break away from the present invention's design, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.
Claims (9)
1. the method based on the semantic Background digging user of notion interest is characterized in that: may further comprise the steps: (A) system server collection user subject data; (B) system server is set up user's theme interest model; (C) system server is set up the semantic Background of notion; (D) system server the more semantic Background of new ideas to upgrade user's theme interesting data.
2. the method based on the semantic Background digging user of notion interest according to claim 1; It is characterized in that: in the said step (A), system server is submitted to search engine to user's query word and in the result who returns page or leaf, is allowed the user select the webpage of being interested in and constitute user's theme interest set.
3. the method based on the semantic Background digging user of notion interest according to claim 1 is characterized in that: in the said step (B), system server is built the concept map of upright reflection user interest jointly through the theme interest set; Said system server made up concept lattice earlier before setting up user's theme interest model.
4. the method based on the semantic Background digging user of notion interest according to claim 3; It is characterized in that: in the said step (C), said system server converts concept lattice into the semantic Background of notion of the semantic relation between can the visual representation webpage.
5. the method based on the semantic Background digging user of notion interest according to claim 1 is characterized in that: in the said step (D), said system server increases or reduces the semantic Background of notion.
6. the method based on the semantic Background digging user of notion interest according to claim 4; It is characterized in that: said system server with the concrete grammar that concept lattice converts the semantic Background of notion of the semantic relation between can the visual representation webpage into is: system server is confirmed key concept earlier and key concept is put into the 0th layer of conceptual backdrop figure; And then utilize what of attribute number of each notion in the concept lattice, the notion in the concept lattice is mapped in the middle of the corresponding level of conceptual backdrop figure.
7. the method based on the semantic Background digging user of notion interest according to claim 3; It is characterized in that: system server utilizes the attribute speech of concept lattice to give the notion layering in the concept lattice; Its layered approach is: the key concept in the concept lattice is inserted among the conceptual backdrop figure, as the 0th layer of conceptual backdrop figure; In the non-core notion of concept lattice, comprise the attribute speech of key concept fully and the node of the attribute speech of higher similarity as the ground floor of conceptual backdrop figure arranged with key concept; In the remaining non-core concept set, comprise the notion of the attribute speech of (N-i+1) individual key concept, as the i layer of conceptual backdrop figure, wherein N representes the number of the attribute speech in the key concept, i ∈ [1, N].
8. the method based on the semantic Background digging user of notion interest according to claim 2; It is characterized in that: system server is collected the webpage that the user selects; Form the interest topic collections of web pages; Again collections of web pages is carried out word segmentation processing, feature extraction; And calculated characteristics weights; Then these characteristics are sorted according to the weights size, the characteristic of selecting to come the front is as the community set that makes up concept lattice.
9. the method based on the semantic Background digging user of notion interest according to claim 8, it is characterized in that: said community set comprises artificial intelligence, machine learning, Knowledge Discovery, agent technology, pattern-recognition, natural language processing, feature extraction, mechanical translation and the representation of knowledge.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110252770A CN102306182A (en) | 2011-08-30 | 2011-08-30 | Method for excavating user interest based on conceptual semantic background image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110252770A CN102306182A (en) | 2011-08-30 | 2011-08-30 | Method for excavating user interest based on conceptual semantic background image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102306182A true CN102306182A (en) | 2012-01-04 |
Family
ID=45380044
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110252770A Pending CN102306182A (en) | 2011-08-30 | 2011-08-30 | Method for excavating user interest based on conceptual semantic background image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102306182A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745405A (en) * | 2014-01-21 | 2014-04-23 | 北京航空航天大学 | Concept map-based destination image measurement method and concept map-based destination image map generation system |
| CN106650940A (en) * | 2016-12-26 | 2017-05-10 | 东软集团股份有限公司 | Field knowledge base establishment method and device |
| CN107169037A (en) * | 2017-04-20 | 2017-09-15 | 河海大学 | A kind of individuation search method of combination sequential dynamic modeling and emotional semantic |
| CN110675217A (en) * | 2019-09-05 | 2020-01-10 | 广州亚美信息科技有限公司 | Personalized background image generation method and device |
| CN111625619A (en) * | 2019-02-28 | 2020-09-04 | 北京沃东天骏信息技术有限公司 | Query omission method and device, computer readable medium and electronic equipment |
| CN112001184A (en) * | 2020-08-14 | 2020-11-27 | 西华大学 | User emotion difference region detection method and system for video bullet screen |
| CN113139389A (en) * | 2021-04-29 | 2021-07-20 | 南宁师范大学 | Graph model semantic query expansion method and device based on dynamic optimization |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706790A (en) * | 2009-09-18 | 2010-05-12 | 浙江大学 | Clustering method of WEB objects in search engine |
-
2011
- 2011-08-30 CN CN201110252770A patent/CN102306182A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706790A (en) * | 2009-09-18 | 2010-05-12 | 浙江大学 | Clustering method of WEB objects in search engine |
Non-Patent Citations (1)
| Title |
|---|
| 彭强强: "基于概念背景图的主题爬行策略研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, 15 May 2011 (2011-05-15), pages 1 - 55 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745405A (en) * | 2014-01-21 | 2014-04-23 | 北京航空航天大学 | Concept map-based destination image measurement method and concept map-based destination image map generation system |
| CN106650940A (en) * | 2016-12-26 | 2017-05-10 | 东软集团股份有限公司 | Field knowledge base establishment method and device |
| CN106650940B (en) * | 2016-12-26 | 2019-01-22 | 东软集团股份有限公司 | A kind of domain knowledge base construction method and device |
| CN107169037A (en) * | 2017-04-20 | 2017-09-15 | 河海大学 | A kind of individuation search method of combination sequential dynamic modeling and emotional semantic |
| CN107169037B (en) * | 2017-04-20 | 2020-06-23 | 河海大学 | Personalized search method combining sequencing dynamic modeling and emotion semantics |
| CN111625619A (en) * | 2019-02-28 | 2020-09-04 | 北京沃东天骏信息技术有限公司 | Query omission method and device, computer readable medium and electronic equipment |
| CN111625619B (en) * | 2019-02-28 | 2024-03-01 | 北京沃东天骏信息技术有限公司 | Query omission method, device, computer readable medium and electronic equipment |
| CN110675217A (en) * | 2019-09-05 | 2020-01-10 | 广州亚美信息科技有限公司 | Personalized background image generation method and device |
| CN112001184A (en) * | 2020-08-14 | 2020-11-27 | 西华大学 | User emotion difference region detection method and system for video bullet screen |
| CN112001184B (en) * | 2020-08-14 | 2022-10-11 | 西华大学 | User emotion difference region detection method and system for video bullet screen |
| CN113139389A (en) * | 2021-04-29 | 2021-07-20 | 南宁师范大学 | Graph model semantic query expansion method and device based on dynamic optimization |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101630314B (en) | Semantic query expansion method based on domain knowledge | |
| CN103491205B (en) | The method for pushing of a kind of correlated resources address based on video search and device | |
| CN102306182A (en) | Method for excavating user interest based on conceptual semantic background image | |
| US8312035B2 (en) | Search engine enhancement using mined implicit links | |
| CN103853831B (en) | Personalized searching realization method based on user interest | |
| CN101399818B (en) | Theme related webpage filtering method and system based on navigation route information | |
| CN105045875B (en) | Personalized search and device | |
| Wang et al. | Ranking user's relevance to a topic through link analysis on web logs | |
| CN103605665A (en) | Keyword based evaluation expert intelligent search and recommendation method | |
| CN102236677A (en) | Question answering system-based information matching method and system | |
| CN105335487A (en) | Agricultural specialist information retrieval system and method on basis of agricultural technology information ontology library | |
| CN103793434A (en) | Content-based image search method and device | |
| CN104794242A (en) | Searching methods | |
| Hsu et al. | Topic-specific crawling on the Web with the measurements of the relevancy context graph | |
| Du et al. | A topic-specific crawling strategy based on semantics similarity | |
| CN103186556A (en) | Method for obtaining and searching structural semantic knowledge and corresponding device | |
| CN105740310A (en) | Automatic answer summarizing method and system for question answering system | |
| CN103020083B (en) | The automatic mining method of demand recognition template, demand recognition methods and corresponding device | |
| CN103064846B (en) | Retrieval device and search method | |
| CN105912637A (en) | Knowledge-based user interest mining method | |
| Jin et al. | Tise: A temporal search engine for web contents | |
| Yan et al. | Research on PageRank and hyperlink-induced topic search in web structure mining | |
| Leung et al. | Constructing concept relation network and its application to personalized web search | |
| CN104794200A (en) | Event publishing and subscribing method supporting fuzzy matching based on ontology | |
| Dou et al. | Low-cost, bottom-up measures for evaluating search result diversification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120104 |