CN105512333A - Product comment theme searching method based on emotional tendency - Google Patents
Product comment theme searching method based on emotional tendency Download PDFInfo
- Publication number
- CN105512333A CN105512333A CN201511003818.XA CN201511003818A CN105512333A CN 105512333 A CN105512333 A CN 105512333A CN 201511003818 A CN201511003818 A CN 201511003818A CN 105512333 A CN105512333 A CN 105512333A
- Authority
- CN
- China
- Prior art keywords
- product
- webpage
- theme
- sentiment orientation
- evaluation
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a product comment theme searching method based on emotional tendency. According to the product comment theme searching method, on the premise of guaranteeing the recall ratio, product comment objects are extracted according to theme models of products, the emotion tendency of product comments is analyzed according to a sentiment word dictionary and output after being sorted, and thus a higher precision ratio is obtained. According to the product comment theme searching method, a ternary set including product objects, product comment phrases and emotion tendency values is adopted for expressing the product comment theme, and emotion tendency of different product comments can be expressed more clearly; besides, collected pages are graded from two aspects, namely relevancy with the theme and the emotion tendency values, the requirement that users are interested in product comments and emotion tendency of the comments is better satisfied, and thus searching accuracy is improved.
Description
Technical field
The present invention relates to a kind of product review subject search method based on Sentiment orientation.
Background technology
In current a lot of search services, there are some for the information search service of the personalization of different user, as the personalized search service based on user behavior analysis, the Query Result returned for the same queries request of different user is also identical to some extent, and namely system can identify the difference in different user individual information demand to a certain extent.Seldom have in subject search current in addition and emotion tendency is carried out to product evaluation, only just search out the correspondence evaluation of corresponding product, but how Search Results but cannot do selection to user helps preferably.Therefore how in the search, not only search out the evaluation of corresponding product, the emotional semantic classification of analysis and text can also be carried out the emotion tendency of product evaluation, and Search Results is sorted become the study hotspot of many scholars in product evaluation subject search field.Nearly ten years to text subject classification research more deep, but to text emotion classification research be also in one less.In the research of text emotion tendency classification, in text, the selection of Sentiment orientation word and extraction are the keys of whole assorting process, and the Sentiment orientation of word differentiates it is the basis of chapter level Sentiment orientation research.
In a lot of theme Meta Search Engine method, Search Results is generally adopted and extracts the method for proper vector, then adopt included angle cosine algorithm, calculate the degree that is consistent of Search Results and theme.But proper vector is discrete, possibly cannot correction search result document, therefore calculating with this will be not accurate enough with the similarity of theme, and the accuracy of Search Results is just affected greatly.
Summary of the invention
The object of the present invention is to provide a kind of product review subject search method based on Sentiment orientation, can under the prerequisite ensureing recall ratio, according to the topic model of product, extract product evaluation object, and according to the emotion tendency that emotion word dictionary analytic product is evaluated, export after sequence, thus obtain higher precision ratio.
For solving the problem, the invention provides a kind of product review subject search method based on Sentiment orientation, comprising:
Set up the topic model of various product, wherein, each topic model comprises multiple product theme;
The seed network address of each product theme according to setting is creeped, the webpage collected creeping processes, extract product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, webpage higher than the threshold value preset retains, and then calculates the Sentiment orientation value of product evaluation of the page;
User is when searching for, and select the product theme that will carry out searching for, then search for according to keyword, Search Results is according to the descending sort of emotion tendency.
Further, in the above-mentioned methods, set up the topic model of various product, comprising:
Topic model takes tlv triple Topic (C, W, V) to represent, forms subject tree structure, wherein: C represents product object; W represents product evaluation phrase; V represents the Sentiment orientation value of product evaluation, and C adopts vector space model (VSM) to represent, uses two tuple C
i(Key
i, Weight
i), wherein, Key
irepresent keyword, Weight
irepresent the weight of keyword, the Sentiment orientation value V of product evaluation is between-1 and 1, and positive number represents to be evaluated the front of product, and value is larger, and Sentiment orientation is higher; Negative number representation is to the unfavorable ratings of product, and value is less, and negative emotion tendency is higher.
Further, in the above-mentioned methods, the seed network address of each product theme according to setting is creeped, comprising:
Several seed network address of creeping is arranged to each product theme, gathers related web page from network.
Further, in the above-mentioned methods, the webpage collected creeping processes, extract product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, the webpage higher than the threshold value preset retains, and then calculate the Sentiment orientation value of product evaluation of the page, comprising:
Extract the text of webpage, and extract the proper vector of the text of webpage;
Calculate the similarity extracting webpage according to the cosine value of the angle of proper vector, remove the webpage repeated;
From remaining webpage, extract evaluation object, calculate the degree of correlation with described product theme according to described evaluation object;
Extract the product evaluation phrase do not abandoned in webpage;
Its Sentiment orientation value is calculated respectively according to dissimilar product evaluation phrase;
Calculate each Sentiment orientation value not abandoning webpage.
Further, in the above-mentioned methods, from remaining webpage, extract evaluation object, calculate the degree of correlation with described product theme according to described evaluation object, comprising:
Participle is carried out to the text in remaining webpage, according to the appearance rule extraction candidate evaluations object of the part-of-speech tagging of participle;
From the angle of recall rate, the part-of-speech rule of evaluation object in the remaining webpage of collection as much as possible;
Webpage adopts the proper vector of product object and relation to represent, the concept of each subclass of product theme is also proper vector, and according to vector space model, the cosine value of two proper vector angles represents their degree of correlation, calculate the degree of correlation of a webpage and theme thus, and record;
According to the relevance threshold of setting, the webpage lower than relevance threshold is abandoned.
Further, in the above-mentioned methods, extract the product evaluation phrase do not abandoned in webpage, comprising:
According to emotion word dictionary, centered by emotion word, be modified into by described evaluation object, degree word and negative word identification and evaluation phrase of assigning to.
Further, in the above-mentioned methods, according to emotion word dictionary, centered by emotion word, assign in the step of identification and evaluation phrase by being modified into of described evaluation object, degree word and negative word, according to emotion word, negative word, degree word and other compositions, product evaluation phrase is divided into 5 classes.
Further, in the above-mentioned methods, calculate each Sentiment orientation value not abandoning webpage, comprising:
By each weights sum not abandoning each evaluation phrase of webpage, judge each emotion tendency not abandoning webpage.
Further, in the above-mentioned methods, in the weights sum by each each evaluation phrase not abandoning webpage,
Pass through formula
calculate each weights sum not abandoning each evaluation phrase of webpage, wherein, W
s(CT) represent one and do not abandon in webpage the Sentiment orientation value evaluating phrase, Document represents a Sentiment orientation value not abandoning webpage, if Document is greater than 0, the text is that front is evaluated; If Document is less than 0, then the text is unfavorable ratings.
Further, in the above-mentioned methods, user is when searching for, and select the product theme that will carry out searching for, then search for according to keyword, Search Results, according to the descending sort of emotion tendency, comprising:
After user entered keyword, search in selected product theme;
The theme selected according to user and key word mate with the webpage collected, and then basis and the height of the correlativity of product theme and the Sentiment orientation value of webpage, demonstrate corresponding webpage according to the descending sort of emotion tendency.
Compared with prior art, the present invention, under the prerequisite ensureing recall ratio, according to the topic model of product, extracts product evaluation object, and according to the emotion tendency that emotion word dictionary analytic product is evaluated, exports, thus obtain higher precision ratio after sequence.The present invention adopts the tlv triple of product object, product evaluation phrase and Sentiment orientation value to represent to the theme of product evaluation, more clearly can express the Sentiment orientation that different product is evaluated.In addition, the page gathered is marked from two aspects, with the degree of correlation of theme and the Sentiment orientation value of product evaluation, meets user better not only to product evaluation, also to the interested needs of Sentiment orientation evaluated, therefore improve the accuracy of search.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the product review subject search method based on Sentiment orientation of one embodiment of the invention.
Embodiment
For enabling above-mentioned purpose of the present invention, feature and advantage become apparent more, and below in conjunction with the drawings and specific embodiments, the present invention is further detailed explanation.
As shown in Figure 1, the invention provides a kind of product review subject search method based on Sentiment orientation, comprising:
Step S1, sets up the topic model of various product, and wherein, each topic model comprises multiple product theme;
Step S2, the seed network address of each product theme according to setting is creeped, the webpage collected creeping processes, extract product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, higher than the webpage reservation of the threshold value preset, and then calculate the Sentiment orientation value of product evaluation of the page;
Step S3, user is when searching for, and select the product theme that will carry out searching for, then search for according to keyword, Search Results is according to the descending sort of emotion tendency.
Preferably, step S1, sets up the topic model of various product, comprising:
Topic model takes tlv triple Topic (C, W, V) to represent, forms subject tree structure, wherein: C represents product object; W represents product evaluation phrase; V represents the Sentiment orientation value of product evaluation, and C adopts vector space model (VSM) to represent, uses two tuple C
i(Key
i, Weight
i), wherein, Key
irepresent keyword, Weight
irepresent the weight of keyword, the Sentiment orientation value V of product evaluation is between-1 and 1, and positive number represents to be evaluated the front of product, and value is larger, and Sentiment orientation is higher; Negative number representation is to the unfavorable ratings of product, and value is less, and negative emotion tendency is higher.
Preferably, the seed network address of each product theme according to setting is creeped, comprising:
Several seed network address of creeping is arranged to each product theme, gathers related web page from network.
Preferably, the webpage collected creeping processes, and extracts product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, the webpage higher than the threshold value preset retains, and then calculate the Sentiment orientation value of product evaluation of the page, comprising:
(1) extract the text of webpage, and extract the proper vector of the text of webpage;
(2) removing duplicate webpages: calculate the similarity extracting webpage according to the cosine value of the angle of proper vector, removes the webpage repeated;
(3) from remaining webpage, extract evaluation object, calculate the degree of correlation with described product theme according to described evaluation object; Preferably, this step comprises: carry out participle to the text in remaining webpage, according to the appearance rule extraction candidate evaluations object of the part-of-speech tagging of participle; From the angle of recall rate, the part-of-speech rule of evaluation object in the remaining webpage of collection as much as possible; Webpage adopts the proper vector of product object and relation to represent, the concept of each subclass of product theme is also proper vector, according to vector space model, the cosine value of two proper vector angles represents their degree of correlation, can calculate the degree of correlation Sim of a webpage and theme thus
j, and record; Simultaneously according to the relevance threshold of setting, the webpage lower than relevance threshold is abandoned;
(4) extract the product evaluation phrase that do not abandon in webpage: according to emotion word dictionary, centered by emotion word, be modified into by described evaluation object, degree word and negative word identification and evaluation phrase of assigning to; Preferably, according to emotion word, negative word, degree word and other compositions, product evaluation phrase is divided into 5 classes;
(5) the Sentiment orientation value of described evaluation phrase is calculated: calculate its Sentiment orientation value respectively according to dissimilar product evaluation phrase;
(6) each Sentiment orientation value not abandoning webpage is calculated: by each weights sum not abandoning each evaluation phrase of webpage, judge each emotion tendency not abandoning webpage; Preferably, formula is passed through
calculate each weights sum not abandoning each evaluation phrase of webpage, wherein, W
s(CT) represent one and do not abandon in webpage the Sentiment orientation value evaluating phrase, Document represents a Sentiment orientation value not abandoning webpage, if Document is greater than 0, the text is that front is evaluated; If Document is less than 0, then the text is unfavorable ratings.
Preferably, step S3, user is when searching for, and select the product theme that will carry out searching for, then search for according to keyword, Search Results, according to the descending sort of emotion tendency, comprising:
After user entered keyword, search in selected product theme;
The theme selected according to user and key word mate with the webpage collected, and then basis and the height of the correlativity of product theme and the Sentiment orientation value of webpage, demonstrate corresponding webpage according to the descending sort of emotion tendency.
In sum, the present invention, under the prerequisite ensureing recall ratio, according to the topic model of product, extracts product evaluation object, and according to the emotion tendency that emotion word dictionary analytic product is evaluated, exports, thus obtain higher precision ratio after sequence.The present invention adopts the tlv triple of product object, product evaluation phrase and Sentiment orientation value to represent to the theme of product evaluation, more clearly can express the Sentiment orientation that different product is evaluated.In addition, the page gathered is marked from two aspects, with the degree of correlation of theme and the Sentiment orientation value of product evaluation, meets user better not only to product evaluation, also to the interested needs of Sentiment orientation evaluated, therefore improve the accuracy of search.
In this instructions, each embodiment adopts the mode of going forward one by one to describe, and what each embodiment stressed is the difference with other embodiments, between each embodiment identical similar portion mutually see.
Professional can also recognize further, in conjunction with unit and the algorithm steps of each example of embodiment disclosed herein description, can realize with electronic hardware, computer software or the combination of the two, in order to the interchangeability of hardware and software is clearly described, generally describe composition and the step of each example in the above description according to function.These functions perform with hardware or software mode actually, depend on application-specific and the design constraint of technical scheme.Professional and technical personnel can use distinct methods to realize described function to each specifically should being used for, but this realization should not thought and exceeds scope of the present invention.
Obviously, those skilled in the art can carry out various change and modification to invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.
Claims (10)
1., based on a product review subject search method for Sentiment orientation, it is characterized in that, comprising:
Set up the topic model of various product, wherein, each topic model comprises multiple product theme;
The seed network address of each product theme according to setting is creeped, the webpage collected creeping processes, extract product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, webpage higher than the threshold value preset retains, and then calculates the Sentiment orientation value of product evaluation of the page;
User is when searching for, and select the product theme that will carry out searching for, then search for according to keyword, Search Results is according to the descending sort of emotion tendency.
2., as claimed in claim 1 based on the product review subject search method of Sentiment orientation, it is characterized in that, set up the topic model of various product, comprising:
Topic model takes tlv triple Topic (C, W, V) to represent, forms subject tree structure, wherein: C represents product object; W represents product evaluation phrase; V represents the Sentiment orientation value of product evaluation, and C adopts vector space model (VSM) to represent, uses two tuple C
i(Key
i, Weight
i), wherein, Key
irepresent keyword, Weight
irepresent the weight of keyword, the Sentiment orientation value V of product evaluation is between-1 and 1, and positive number represents to be evaluated the front of product, and value is larger, and Sentiment orientation is higher; Negative number representation is to the unfavorable ratings of product, and value is less, and negative emotion tendency is higher.
3. as claimed in claim 1 based on the product review subject search method of Sentiment orientation, it is characterized in that, the seed network address of each product theme according to setting creeped, comprising:
Several seed network address of creeping is arranged to each product theme, gathers related web page from network.
4. as claimed in claim 1 based on the product review subject search method of Sentiment orientation, it is characterized in that, the webpage collected creeping processes, extract product object and product evaluation phrase, according to product topic model, calculate the degree of correlation of webpage and theme, the webpage higher than the threshold value preset retains, and then calculate the Sentiment orientation value of product evaluation of the page, comprising:
Extract the text of webpage, and extract the proper vector of the text of webpage;
Calculate the similarity extracting webpage according to the cosine value of the angle of proper vector, remove the webpage repeated;
From remaining webpage, extract evaluation object, calculate the degree of correlation with described product theme according to described evaluation object;
Extract the product evaluation phrase do not abandoned in webpage;
Its Sentiment orientation value is calculated respectively according to dissimilar product evaluation phrase;
Calculate each Sentiment orientation value not abandoning webpage.
5. as claimed in claim 4 based on the product review subject search method of Sentiment orientation, it is characterized in that, from remaining webpage, extract evaluation object, calculate the degree of correlation with described product theme according to described evaluation object, comprising:
Participle is carried out to the text in remaining webpage, according to the appearance rule extraction candidate evaluations object of the part-of-speech tagging of participle;
From the angle of recall rate, the part-of-speech rule of evaluation object in the remaining webpage of collection as much as possible;
Webpage adopts the proper vector of product object and relation to represent, the concept of each subclass of product theme is also proper vector, and according to vector space model, the cosine value of two proper vector angles represents their degree of correlation, calculate the degree of correlation of a webpage and theme thus, and record;
According to the relevance threshold of setting, the webpage lower than relevance threshold is abandoned.
6. as claimed in claim 4 based on the product review subject search method of Sentiment orientation, it is characterized in that, extract the product evaluation phrase do not abandoned in webpage, comprising:
According to emotion word dictionary, centered by emotion word, be modified into by described evaluation object, degree word and negative word identification and evaluation phrase of assigning to.
7. as claimed in claim 6 based on the product review subject search method of Sentiment orientation, it is characterized in that, according to emotion word dictionary, centered by emotion word, assign in the step of identification and evaluation phrase by being modified into of described evaluation object, degree word and negative word, according to emotion word, negative word, degree word and other compositions, product evaluation phrase is divided into 5 classes.
8., as claimed in claim 4 based on the product review subject search method of Sentiment orientation, it is characterized in that, calculate each Sentiment orientation value not abandoning webpage, comprising:
By each weights sum not abandoning each evaluation phrase of webpage, judge each emotion tendency not abandoning webpage.
9., as claimed in claim 8 based on the product review subject search method of Sentiment orientation, it is characterized in that, in the weights sum by each each evaluation phrase not abandoning webpage,
Pass through formula
calculate each weights sum not abandoning each evaluation phrase of webpage, wherein, W
s(CT) represent one and do not abandon in webpage the Sentiment orientation value evaluating phrase, Document represents a Sentiment orientation value not abandoning webpage, if Document is greater than 0, the text is that front is evaluated; If Document is less than 0, then the text is unfavorable ratings.
10., as claimed in claim 1 based on the product review subject search method of Sentiment orientation, it is characterized in that, user is when searching for, selection will carry out the product theme searched for, then search for according to keyword, Search Results, according to the descending sort of emotion tendency, comprising:
After user entered keyword, search in selected product theme;
The theme selected according to user and key word mate with the webpage collected, and then basis and the height of the correlativity of product theme and the Sentiment orientation value of webpage, demonstrate corresponding webpage according to the descending sort of emotion tendency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511003818.XA CN105512333A (en) | 2015-12-28 | 2015-12-28 | Product comment theme searching method based on emotional tendency |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511003818.XA CN105512333A (en) | 2015-12-28 | 2015-12-28 | Product comment theme searching method based on emotional tendency |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105512333A true CN105512333A (en) | 2016-04-20 |
Family
ID=55720313
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511003818.XA Pending CN105512333A (en) | 2015-12-28 | 2015-12-28 | Product comment theme searching method based on emotional tendency |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105512333A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021562A (en) * | 2016-05-31 | 2016-10-12 | 北京京拍档科技有限公司 | Method for recommending E-commerce platform based on theme relevance |
| CN106469145A (en) * | 2016-09-30 | 2017-03-01 | 中科鼎富（北京）科技发展有限公司 | Text emotion analysis method and device |
| CN107220352A (en) * | 2017-05-31 | 2017-09-29 | 北京百度网讯科技有限公司 | The method and apparatus that comment collection of illustrative plates is built based on artificial intelligence |
| CN107305574A (en) * | 2016-04-25 | 2017-10-31 | 百度在线网络技术（北京）有限公司 | Object search method and device |
| CN107767195A (en) * | 2016-08-16 | 2018-03-06 | 阿里巴巴集团控股有限公司 | The display systems and displaying of description information, generation method and electronic equipment |
| CN110096694A (en) * | 2018-01-30 | 2019-08-06 | 北京京东尚科信息技术有限公司 | Information generating method and device based on natural language processing |
| CN110674415A (en) * | 2019-09-20 | 2020-01-10 | 北京浪潮数据技术有限公司 | Information display method and device and server |
| CN113254777A (en) * | 2021-06-07 | 2021-08-13 | 武汉卓尔数字传媒科技有限公司 | Information recommendation method and device, electronic equipment and storage medium |
| CN115795040A (en) * | 2023-02-10 | 2023-03-14 | 成都桉尼维尔信息科技有限公司 | User portrait analysis method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102236722A (en) * | 2011-08-17 | 2011-11-09 | 广州索答信息科技有限公司 | Method and system for generating user comment summaries based on triples |
| CN103823893A (en) * | 2014-03-11 | 2014-05-28 | 北京大学 | User comment-based product search method and system |
| CN104484815A (en) * | 2014-12-18 | 2015-04-01 | 刘耀强 | Product-oriented emotion analysis method and system based on fuzzy body |
-
2015
- 2015-12-28 CN CN201511003818.XA patent/CN105512333A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102236722A (en) * | 2011-08-17 | 2011-11-09 | 广州索答信息科技有限公司 | Method and system for generating user comment summaries based on triples |
| CN103823893A (en) * | 2014-03-11 | 2014-05-28 | 北京大学 | User comment-based product search method and system |
| CN104484815A (en) * | 2014-12-18 | 2015-04-01 | 刘耀强 | Product-oriented emotion analysis method and system based on fuzzy body |
Non-Patent Citations (1)
| Title |
|---|
| 杜霖: "《基于句子结构化特征的情感倾向分析》", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107305574A (en) * | 2016-04-25 | 2017-10-31 | 百度在线网络技术（北京）有限公司 | Object search method and device |
| CN106021562B (en) * | 2016-05-31 | 2019-05-24 | 北京京拍档科技有限公司 | For electric business platform based on the relevant recommended method of theme |
| CN106021562A (en) * | 2016-05-31 | 2016-10-12 | 北京京拍档科技有限公司 | Method for recommending E-commerce platform based on theme relevance |
| CN107767195A (en) * | 2016-08-16 | 2018-03-06 | 阿里巴巴集团控股有限公司 | The display systems and displaying of description information, generation method and electronic equipment |
| CN106469145A (en) * | 2016-09-30 | 2017-03-01 | 中科鼎富（北京）科技发展有限公司 | Text emotion analysis method and device |
| CN107220352B (en) * | 2017-05-31 | 2020-12-08 | 北京百度网讯科技有限公司 | Method and device for constructing comment map based on artificial intelligence |
| US10642938B2 (en) | 2017-05-31 | 2020-05-05 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Artificial intelligence based method and apparatus for constructing comment graph |
| CN107220352A (en) * | 2017-05-31 | 2017-09-29 | 北京百度网讯科技有限公司 | The method and apparatus that comment collection of illustrative plates is built based on artificial intelligence |
| CN110096694A (en) * | 2018-01-30 | 2019-08-06 | 北京京东尚科信息技术有限公司 | Information generating method and device based on natural language processing |
| CN110674415A (en) * | 2019-09-20 | 2020-01-10 | 北京浪潮数据技术有限公司 | Information display method and device and server |
| CN110674415B (en) * | 2019-09-20 | 2022-06-17 | 北京浪潮数据技术有限公司 | Information display method and device and server |
| CN113254777A (en) * | 2021-06-07 | 2021-08-13 | 武汉卓尔数字传媒科技有限公司 | Information recommendation method and device, electronic equipment and storage medium |
| CN115795040A (en) * | 2023-02-10 | 2023-03-14 | 成都桉尼维尔信息科技有限公司 | User portrait analysis method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512333A (en) | Product comment theme searching method based on emotional tendency | |
| CN106649260B (en) | Product characteristic structure tree construction method based on comment text mining | |
| CN103544255B (en) | Text semantic relativity based network public opinion information analysis method | |
| KR102080362B1 (en) | Query expansion | |
| CN101320375B (en) | Digital book search method based on user click action | |
| CN106599054B (en) | Method and system for classifying and pushing questions | |
| CN103226578B (en) | Towards the website identification of medical domain and the method for webpage disaggregated classification | |
| CN106156372B (en) | A kind of classification method and device of internet site | |
| CN108197117A (en) | A kind of Chinese text keyword extracting method based on document subject matter structure with semanteme | |
| CN103838833A (en) | Full-text retrieval system based on semantic analysis of relevant words | |
| CN106339502A (en) | Modeling recommendation method based on user behavior data fragmentation cluster | |
| CN109271477A (en) | A kind of method and system by internet building taxonomy library | |
| CN103455487B (en) | The extracting method and device of a kind of search term | |
| JP2005085285A5 (en) | ||
| CN104778209A (en) | Opinion mining method for ten-million-scale news comments | |
| CN104866572A (en) | Method for clustering network-based short texts | |
| CN104866554B (en) | A kind of individuation search method and system based on socialization mark | |
| CN101127042A (en) | Sensibility classification method based on language model | |
| TW201214169A (en) | Recognition of target words using designated characteristic values | |
| CN101944099A (en) | Method for automatically classifying text documents by utilizing body | |
| CN104317834A (en) | Cross-media sorting method based on deep neural network | |
| CN106126502A (en) | A kind of emotional semantic classification system and method based on support vector machine | |
| Sabuna et al. | Summarizing Indonesian text automatically by using sentence scoring and decision tree | |
| CN103049470A (en) | Opinion retrieval method based on emotional relevancy | |
| CN106484797A (en) | Accident summary abstracting method based on sparse study |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160420 |
|
| WD01 | Invention patent application deemed withdrawn after publication |