CN107704556A - A kind of sentiment analysis method and system in automobile industry subdivision field - Google Patents
A kind of sentiment analysis method and system in automobile industry subdivision field Download PDFInfo
- Publication number
- CN107704556A CN107704556A CN201710896269.6A CN201710896269A CN107704556A CN 107704556 A CN107704556 A CN 107704556A CN 201710896269 A CN201710896269 A CN 201710896269A CN 107704556 A CN107704556 A CN 107704556A
- Authority
- CN
- China
- Prior art keywords
- dictionary
- automobile industry
- user
- original content
- subdivision field
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/374—Thesaurus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/355—Class or cluster creation or modification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Machine Translation (AREA)
- Document Processing Apparatus (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of sentiment analysis method and system in automobile industry subdivision field.This method includes：Obtain automobile industry dictionary；According to user's original content of the automobile industry as training material and the automobile industry dictionary training emotion classifiers；The emotion score value of user's original content of automobile industry to be evaluated is calculated according to the automobile industry dictionary and/or the emotion classifiers.The present invention is by building evaluation entity and sentiment dictionary under automobile industry subdivision field, respective sentiment dictionary, which is based on, for different subdivision fields carries out emotion score value calculating, then carry out the calculating of emotion score value in the case of miss by way of machine learning again for sentiment dictionary, both solved the defects of sentiment dictionary coverage, and alleviated impure the problem of causing the grader degree of accuracy to decline due to language material to a certain extent.
Description
Technical field
The present invention relates to big data field, more particularly to a kind of automobile industry to segment the sentiment analysis method in field and be
System.
Background technology
With the fast development of internet, it is more and more to segment the website in field, such as the service such as automobile, food and drink, lodging
Website.Many users, along with the practical experience of oneself, can deliver the sight of oneself in subdivision field for the service of oneself
Point.The feedback of user has great importance to product or ISP, particularly follows the trail of user to product or service
Public sentiment trend and to product or the follow-up improvement of service.
In auto Life, people for buying car, with being run into during car the problem of, will also tend in microblogging, forum, patch
Etc. public arena deliver the opinion of oneself, so for Automobile Enterprises, pass through automobile public sentiment and monitor the moment and pay close attention to user
Evaluation and feedback to product, not only it can be intervened timely processing, Er Qieke in advance to the focus event being likely to occur
To collect the suggestions on Optimization of many products., often can be by automobile industry related web site in automobile public sentiment system
Or the speech that targeted customer delivers in forum is analyzed in real time, the positive negative emotion that user goes out expressed by product is obtained
Tendency, so as to reach the public sentiment monitoring to Related product.
It is positive and negative in Main Basiss sentiment dictionary in the prior art, such as basic sentiment dictionary and automobile industry sentiment dictionary
The emotion such as face emotion word and adjective differentiates, or carries out just negative emotional semantic classification to text based on the method for machine learning,
So as to obtain the Affective Evaluation of user's original content.
Prior art can calculate emotion score value, but there are the following problems：Have to the range of automobile industry sentiment dictionary
High requirement, if the coverage of dictionary is smaller, that is, just negative emotion word is much smaller than the emotion word number of necessary being
Amount, is so easy to fail to judge to the emotion word in model, so as to reduce the accuracy of model Sentiment orientation judgement；It is same in dictionary
Individual emotion word gives expression to antipodal Sentiment orientation sometimes under different context, can not determine the Affective Evaluation of emotion word；
Prior art very broadly calculates a score value for every model, can not navigate to the Affective Evaluation to details.
The content of the invention
In view of this, the present invention provides a kind of sentiment analysis method and system in automobile industry subdivision field, and its purpose exists
In realizing the sentiment analysis under automobile industry subdivision field.
The invention provides a kind of sentiment analysis method in automobile industry subdivision field, this method includes：
Obtain automobile industry dictionary；
According to user's original content of the automobile industry as training material and the automobile industry dictionary training emotion
Grader；
The user that automobile industry to be evaluated is calculated according to the automobile industry dictionary and/or the emotion classifiers is original
The emotion score value of content.
Preferably, the automobile industry dictionary includes the entity dictionary in subdivision field and the sentiment dictionary in subdivision field；Its
In, the entity dictionary in the subdivision field is run after fame part of speech dictionary, and the sentiment dictionary in the subdivision field is adjectival dictionary.
Preferably, the basis is used as user's original content of the automobile industry of training material and the automobile industry word
Allusion quotation training emotion classifiers include：
The user's original content institute of automobile industry for being used as training material is judged according to the entity dictionary in the subdivision field
Belong to subdivision field；
Generation is used as the term vector expression of user's original content of the automobile industry of training material；
The training data of tape label is generated according to comment most satisfied in automobile public praise data and most unsatisfied comment；
The sentence vector of the user's original content for the automobile industry for being used as training material is calculated, and trains the subdivision field
Emotion classifiers.
Preferably, it is described that garage to be evaluated is calculated according to the automobile industry dictionary and/or the emotion classifiers
The emotion score value of user's original content of industry includes：
The subdivision according to belonging to the entity dictionary in the subdivision field determines user's original content of automobile industry to be evaluated
Field；
Determining that user's original content of the automobile industry to be evaluated is deposited according to the sentiment dictionary in the subdivision field
In sentiment dictionary in the case of contained adjectival word, according to the sentiment dictionary, negative word dictionary and journey in the subdivision field
Spend the emotion score value that adverbial word dictionary calculates user's original content of automobile industry to be evaluated；
User's original content of the automobile industry to be evaluated is being determined not according to the sentiment dictionary in the subdivision field
Exist in sentiment dictionary in the case of contained adjectival word, calculated according to the emotion classifiers in the subdivision field described to be evaluated
The emotion score value of user's original content of the automobile industry of valency.
Preferably, the generation is used as training the term vector expression of user's original content of the automobile industry of material to include：
It is used as the term vector table of user's original content of the automobile industry of training material according to the generation of word2vector models
Up to including.
Preferably, the emotion classifiers for training the subdivision field include：
The emotion classifiers in the subdivision field are trained according to gradient boosted tree.
The invention provides a kind of sentiment analysis system in automobile industry subdivision field, the system includes：
Acquisition module, for obtaining automobile industry dictionary；
Training module, for user's original content according to the automobile industry for being used as training material and the automobile industry
Dictionary training emotion classifiers；
Computing module, for calculating automobile to be evaluated according to the automobile industry dictionary and/or the emotion classifiers
The emotion score value of user's original content of industry.
Preferably, the automobile industry dictionary includes the entity dictionary in subdivision field and the sentiment dictionary in subdivision field；Its
In, the entity dictionary in the subdivision field is run after fame part of speech dictionary, and the sentiment dictionary in the subdivision field is adjectival dictionary.
Preferably, the training module is additionally operable to be judged to be used as training material according to the entity dictionary in the subdivision field
Subdivision field belonging to user's original content of automobile industry；Generation is used as user's original content of the automobile industry of training material
Term vector is expressed；The training number of tape label is generated according to comment most satisfied in automobile public praise data and most unsatisfied comment
According to；The sentence vector of the user's original content for the automobile industry for being used as training material is calculated, and trains the emotion in the subdivision field
Grader.
Preferably, the computing module is additionally operable to determine garage to be evaluated according to the entity dictionary in the subdivision field
Subdivision field belonging to user's original content of industry；The automobile to be evaluated is being determined according to the sentiment dictionary in the subdivision field
User's original content of industry is present in sentiment dictionary in the case of contained adjectival word, according to the emotion in the subdivision field
Dictionary, negative word dictionary and degree adverb dictionary calculate the emotion score value of user's original content of automobile industry to be evaluated；
Emotion word is not present in the user's original content for determining the automobile industry to be evaluated according to the sentiment dictionary in the subdivision field
In allusion quotation in the case of contained adjectival word, the garage to be evaluated is calculated according to the emotion classifiers in the subdivision field
The emotion score value of user's original content of industry.
Preferably, the training module is additionally operable to the garage for being used as training material according to the generation of word2vector models
The term vector expression of user's original content of industry includes.
Preferably, the training module is additionally operable to train the emotion classifiers in the subdivision field according to gradient boosted tree.
The present invention is led by building evaluation entity and sentiment dictionary under automobile industry subdivision field for different subdivisions
Domain is based on respective sentiment dictionary and carries out emotion score value calculating, not only can greatly expand automobile industry sentiment dictionary, and
Industry sentiment dictionary is classified, solves the ambiguity problem of emotion word under conventional different context, then for sentiment dictionary
Carry out the calculating of emotion score value in the case of miss by way of machine learning again, both solve sentiment dictionary coverage
The defects of, and impure the problem of causing the grader degree of accuracy to decline due to language material is alleviated to a certain extent, pass through subdivision
The sentiment dictionary in field and the method for machine learning realize the sentiment analysis under automobile industry subdivision field jointly.
Brief description of the drawings
Accompanying drawing is for providing a further understanding of the present invention, and a part for constitution instruction, with following tool
Body embodiment is used to explain the present invention together, but is not construed as limiting the invention.In the accompanying drawings：
Fig. 1 is dictionary creation schematic flow sheet provided by the invention；
Fig. 2 is the sentiment analysis method schematic diagram in automobile industry subdivision field provided by the invention；
Fig. 3 is emotion classifiers training flow chart provided by the invention；
Fig. 4 is the detail flowchart of the sentiment analysis in automobile industry subdivision field provided by the invention；
Fig. 5 is the sentiment analysis system schematic in automobile industry subdivision field provided by the invention.
Embodiment
The embodiment of the present invention is described in detail below in conjunction with accompanying drawing.It should be appreciated that this place is retouched
The embodiment stated is merely to illustrate and explain the present invention, and is not intended to limit the invention.
As shown in figure 1, the flow chart of structure noun dictionary provided by the invention and adjective dictionary, is specifically included：
Step 105, the public praise language material under certain car website subdivision field, such as the public praise language under 8 subdivision fields are captured
Material, subdivision field can be such as space, oil consumption, power, manipulation, outward appearance, interior trim, cost performance, comfortableness；Public praise language material can be with
Captured according to the keyword in subdivision field.
Step 110, language material segmented, remove stop words, go vehicle, going low-frequency word etc. to pre-process；Participle can use
Some segmentation methods, such as the segmentation methods based on string matching, the segmentation methods based on understanding, the participle based on statistics are calculated
Method etc.；Stop words such as some onomatopoeias, conjunction etc., these words are not acted on for the sentiment analysis of language material, can removed；Car
Type can determine according to database, vehicle can be assert according to machine learning, so as to remove；Low-frequency word refer to life in compared with
The word run into less, or some classical Chinese vocabulary etc..
Step 115, legal name part of speech word is extracted according to screening rule, and counts word frequency；Name part of speech word is noun,
Engine, power, clutch, gear etc. are common are in automobile industry；Can be according to multiple users when statistics word frequency
Original content is to calculate, such as appearance is how many inferior per K word.
Step 120, the mutual information of calculating field and name part of speech word, the word of top200 under each subdivision field is chosen；Mutual information,
It is two complementary measurements of stochastic variable, the value shows that more greatly both correlations are stronger, is led here by each subdivision
Original content under domain is segmented and part-of-speech tagging, reserved name part of speech word, then calculates these part of speech words and is led with each subdivision
The mutual information in domain；
Step 125, the entity dictionary under subdivision field, i.e. noun dictionary are generated by artificial screening；Included in dictionary
Word is such as engine, power, clutch, gear；
Step 130, legal adjectival word is extracted according to screening rule, and counts word frequency；Adjectival word is
Adjective, such as good, poor, meat, powerful etc.；It can be calculated when counting word frequency according to multiple user's original contents, such as
Occur per K word how many inferior.
Step 135, the mutual information of calculating field and adjectival word, the word of top200 under each subdivision field is chosen
Step 140, the sentiment dictionary under subdivision field, i.e. adjective dictionary are generated by artificial screening.Included in dictionary
Word such as good, poor, meat, it is powerful.
Based on constructed name part of speech dictionary and adjectival dictionary, the invention provides a kind of automobile industry to segment field
Sentiment analysis method, as shown in Fig. 2 specifically including：
Step 205, automobile industry dictionary is obtained；Automobile industry dictionary can be included under the automobile subdivision field of above-mentioned structure
Name part of speech dictionary and adjectival dictionary；
Step 210, according to user's original content of the automobile industry as training material and automobile industry dictionary training
Emotion classifiers；User's original content as training material can pass through artificial screening, to improve the training of emotion classifiers
Effect；
Step 215, the user that automobile industry to be evaluated is calculated according to automobile industry dictionary and/or emotion classifiers is original
The emotion score value of content.In the step, if having hit adjective dictionary, emotion point directly can be calculated according to adjective
Value, if not hitting, can calculate emotion score value according to emotion classifiers.
For automobile industry automobile original content to be analyzed, it is thus necessary to determine that subdivision field belonging to the text, be then based on
The method of dictionary judges whether to hit the emotion word under the subdivision field, if do not hit, using the side of emotion classifiers
Method carries out sentiment analysis, can so make up sentiment dictionary it is born possessed by can not include all emotion words forever and bring
Problem of failing to judge.
Fig. 3 is emotion classifiers training flow chart provided by the invention, is specifically included：
Step 305, by the entity dictionary under subdivision field, subdivision field belonging to text is judged；
Step 310, the term vector table of automobile industry user's original content of training material is used as based on vector model generation
Reach, such as based on word2vector models；
Step 315, the training data of tape label is generated by most satisfied in automobile public praise data and most unsatisfied comment；
Most satisfied and most unsatisfied mark is included in car website original content, retains during crawler capturing；
The data of tape label are used for training emotion classifiers；
Step 320, the vector space expression of text is calculated, based on the emotional semantic classification under gradient boosted tree training subdivision field
Device.The space vector expression of text can be understood as the space vector expression of sentence, be the term vector based on contained word in sentence
Expression generation.Basis is most full before sentence vector constitutes the sample for training, and the sample label for being used to train is exactly
(satisfied is 1, discontented to mean 0) for meaning and the label data of least satisfied generation.
The emotion classifiers that above-mentioned training obtains can be used in the case of no hit adjective dictionary, with emotion point
Class device evaluates the emotion of user.
Fig. 4 is the detail flowchart of the sentiment analysis in automobile industry subdivision field provided by the invention, including：
Step 405, automobile industry user original content to be analyzed is obtained；
Step 410, based on the entity dictionary under subdivision field, the affiliated automobile industry subdivision neck of user's original content is judged
Domain；
Step 415, judge whether user's original content hits the adjective in adjective dictionary, if execution step
420, otherwise perform step 425；
Step 420, based on the sentiment dictionary under subdivision field, negative word dictionary and degree adverb dictionary, it is former to calculate user
Create the emotion score value of content；Negative word dictionary includes：No, it is non-, vocabulary is less waited, for judging whether to invert feeling polarities；Journey
Degree adverbial word dictionary includes：Very, very, vocabulary is somewhat waited, for judging degree；Negative word dictionary and Chengdu adverbial word dictionary can also
Pre-establish.Negative word dictionary and degree adverb dictionary belong to universaling dictionary, are obtained according to some online disclosed dictionarys, after
Continuous can be changed according to effect additions and deletions is looked into.
Step 425, based on the emotion classifiers under subdivision field, the emotion score value of calculating user's original content.If not
Emotion score value can be got based on dictionary, the vector table that content is generated based on term vector expression before is reached, and is then put it into point
The calculating of emotion score value is carried out in class device.
Based on above flow, even in the situation of no hit sentiment dictionary, the present invention can also be to user's original content
Carry out emotion scoring.
The preferred embodiment of the present invention is described in detail above in association with accompanying drawing, still, the present invention is not limited to above-mentioned reality
The detail in mode is applied, in the range of the technology design of the present invention, a variety of letters can be carried out to technical scheme
Monotropic type, these simple variants belong to protection scope of the present invention.
It is further to note that each particular technique feature described in above-mentioned embodiment, in not lance
In the case of shield, it can be combined by any suitable means.In order to avoid unnecessary repetition, the present invention to it is various can
The combination of energy no longer separately illustrates.
In addition, various embodiments of the present invention can be combined randomly, as long as it is without prejudice to originally
The thought of invention, it should equally be considered as content disclosed in this invention.
Claims (12)
- A kind of 1. sentiment analysis method in automobile industry subdivision field, it is characterised in that this method includes：Obtain automobile industry dictionary；According to user's original content of the automobile industry as training material and the automobile industry dictionary training emotional semantic classification Device；User's original content of automobile industry to be evaluated is calculated according to the automobile industry dictionary and/or the emotion classifiers Emotion score value.
- 2. according to the method for claim 1, it is characterised in that the automobile industry dictionary includes the entity word in subdivision field Allusion quotation and the sentiment dictionary in subdivision field；Wherein, the entity dictionary in the subdivision field is run after fame part of speech dictionary, the subdivision field Sentiment dictionary is adjectival dictionary.
- 3. according to the method for claim 2, it is characterised in that the basis is used as the user of the automobile industry of training material Original content and the automobile industry dictionary training emotion classifiers include：It is thin belonging to the user's original content for judging the automobile industry as training material according to the entity dictionary in the subdivision field Divide field；Generation is used as the term vector expression of user's original content of the automobile industry of training material；The training data of tape label is generated according to comment most satisfied in automobile public praise data and most unsatisfied comment；The sentence vector of the user's original content for the automobile industry for being used as training material is calculated, and trains the emotion in the subdivision field Grader.
- 4. according to the method for claim 3, it is characterised in that described according to the automobile industry dictionary and/or the feelings The emotion score value of user's original content of sense classifier calculated automobile industry to be evaluated includes：Field is segmented according to belonging to the entity dictionary in the subdivision field determines user's original content of automobile industry to be evaluated；Determining that user's original content of the automobile industry to be evaluated has feelings according to the sentiment dictionary in the subdivision field Feel in dictionary in the case of contained adjectival word, according to the sentiment dictionary, negative word dictionary and degree pair in the subdivision field Word dictionary calculates the emotion score value of user's original content of automobile industry to be evaluated；Determining that user's original content of the automobile industry to be evaluated is not present according to the sentiment dictionary in the subdivision field In sentiment dictionary in the case of contained adjectival word, calculated according to the emotion classifiers in the subdivision field described to be evaluated The emotion score value of user's original content of automobile industry.
- 5. according to the method for claim 3, it is characterised in that the generation is used as the user of the automobile industry of training material The term vector expression of original content includes：The term vector for being used as user's original content of the automobile industry of training material according to the generation of word2vector models expresses bag Include.
- 6. according to the method for claim 3, it is characterised in that the emotion classifiers for training the subdivision field include：The emotion classifiers in the subdivision field are trained according to gradient boosted tree.
- 7. a kind of sentiment analysis system in automobile industry subdivision field, it is characterised in that the system includes：Acquisition module, for obtaining automobile industry dictionary；Training module, for user's original content according to the automobile industry for being used as training material and the automobile industry dictionary Train emotion classifiers；Computing module, for calculating automobile industry to be evaluated according to the automobile industry dictionary and/or the emotion classifiers User's original content emotion score value.
- 8. system according to claim 7, it is characterised in that the automobile industry dictionary includes the entity word in subdivision field Allusion quotation and the sentiment dictionary in subdivision field；Wherein, the entity dictionary in the subdivision field is run after fame part of speech dictionary, the subdivision field Sentiment dictionary is adjectival dictionary.
- 9. system according to claim 8, it is characterised in that the training module is additionally operable to according to the subdivision field Entity dictionary judges to be used as subdivision field belonging to user's original content of the automobile industry of training material；Generation is used as training material Automobile industry user's original content term vector expression；It is according to comment most satisfied in automobile public praise data and least satisfied Comment generation tape label training data；Calculate be used as training material automobile industry user's original content sentence to Amount, and train the emotion classifiers in the subdivision field.
- 10. system according to claim 9, it is characterised in that the computing module is additionally operable to according to the subdivision field Entity dictionary determine subdivision field belonging to user's original content of automobile industry to be evaluated；According to the subdivision field Sentiment dictionary determines that user's original content of the automobile industry to be evaluated has contained adjectival word in sentiment dictionary In the case of, garage to be evaluated is calculated according to the sentiment dictionary, negative word dictionary and degree adverb dictionary in the subdivision field The emotion score value of user's original content of industry；The garage to be evaluated is being determined according to the sentiment dictionary in the subdivision field User's original content of industry is not present in sentiment dictionary in the case of contained adjectival word, according to the emotion in the subdivision field The emotion score value of user's original content of automobile industry to be evaluated described in classifier calculated.
- 11. system according to claim 9, it is characterised in that the training module is additionally operable to according to word2vector moulds The term vector expression that type generation is used as user's original content of the automobile industry of training material includes.
- 12. system according to claim 9, it is characterised in that the training module is additionally operable to be instructed according to gradient boosted tree Practice the emotion classifiers in the subdivision field.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710896269.6A CN107704556B (en) | 2017-09-28 | 2017-09-28 | Emotion analysis method and system for automobile industry subdivision field |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710896269.6A CN107704556B (en) | 2017-09-28 | 2017-09-28 | Emotion analysis method and system for automobile industry subdivision field |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107704556A true CN107704556A (en) | 2018-02-16 |
| CN107704556B CN107704556B (en) | 2020-04-24 |
Family
ID=61175209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710896269.6A Active CN107704556B (en) | 2017-09-28 | 2017-09-28 | Emotion analysis method and system for automobile industry subdivision field |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107704556B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109408809A (en) * | 2018-09-25 | 2019-03-01 | 天津大学 | A kind of sentiment analysis method for automobile product comment based on term vector |
| CN109635287A (en) * | 2018-11-26 | 2019-04-16 | 平安科技（深圳）有限公司 | Method, apparatus, computer equipment and the storage medium of policy dynamics analysis |
| CN109871533A (en) * | 2019-01-04 | 2019-06-11 | 北京车慧科技有限公司 | A kind of corpus processing system based on corpus field |
| CN110442728A (en) * | 2019-06-28 | 2019-11-12 | 天津大学 | Sentiment dictionary construction method based on word2vec automobile product field |
| CN110728131A (en) * | 2018-06-29 | 2020-01-24 | 北京京东尚科信息技术有限公司 | Method and device for analyzing text attribute |
| CN111144929A (en) * | 2019-12-04 | 2020-05-12 | 天津大学 | Comment object and word combined extraction method for automobile industry user generated content |
| CN111611461A (en) * | 2019-05-14 | 2020-09-01 | 北京精准沟通传媒科技股份有限公司 | Data processing method and device |
| CN111767399A (en) * | 2020-06-30 | 2020-10-13 | 平安国际智慧城市科技股份有限公司 | Emotion classifier construction method, device, equipment and medium based on unbalanced text set |
| CN114547167A (en) * | 2022-01-27 | 2022-05-27 | 启明信息技术股份有限公司 | Automobile public opinion sentiment analysis method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103995853A (en) * | 2014-05-12 | 2014-08-20 | 中国科学院计算技术研究所 | Multi-language emotional data processing and classifying method and system based on key sentences |
-
2017
- 2017-09-28 CN CN201710896269.6A patent/CN107704556B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103995853A (en) * | 2014-05-12 | 2014-08-20 | 中国科学院计算技术研究所 | Multi-language emotional data processing and classifying method and system based on key sentences |
Non-Patent Citations (3)
| Title |
|---|
| 库斯曼特等: "《高级数据库营销 互联网时代持续提高客户终身价值的全新方法与实践》", 30 October 2014, 企业管理出版社 * |
| 王庆林等: "《气候变化领域本体手册》", 31 May 2015, 北京理工大学出版社 * |
| 黄鹤: ""面向汽车在线评论的情感分类研究与应用"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110728131A (en) * | 2018-06-29 | 2020-01-24 | 北京京东尚科信息技术有限公司 | Method and device for analyzing text attribute |
| CN109408809A (en) * | 2018-09-25 | 2019-03-01 | 天津大学 | A kind of sentiment analysis method for automobile product comment based on term vector |
| CN109635287A (en) * | 2018-11-26 | 2019-04-16 | 平安科技（深圳）有限公司 | Method, apparatus, computer equipment and the storage medium of policy dynamics analysis |
| CN109871533A (en) * | 2019-01-04 | 2019-06-11 | 北京车慧科技有限公司 | A kind of corpus processing system based on corpus field |
| CN111611461A (en) * | 2019-05-14 | 2020-09-01 | 北京精准沟通传媒科技股份有限公司 | Data processing method and device |
| CN110442728A (en) * | 2019-06-28 | 2019-11-12 | 天津大学 | Sentiment dictionary construction method based on word2vec automobile product field |
| CN111144929A (en) * | 2019-12-04 | 2020-05-12 | 天津大学 | Comment object and word combined extraction method for automobile industry user generated content |
| CN111767399A (en) * | 2020-06-30 | 2020-10-13 | 平安国际智慧城市科技股份有限公司 | Emotion classifier construction method, device, equipment and medium based on unbalanced text set |
| CN114547167A (en) * | 2022-01-27 | 2022-05-27 | 启明信息技术股份有限公司 | Automobile public opinion sentiment analysis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107704556B (en) | 2020-04-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107704556A (en) | A kind of sentiment analysis method and system in automobile industry subdivision field | |
| CN108363753B (en) | Comment text emotion classification model training and emotion classification method, device and equipment | |
| Cao et al. | Web-based traffic sentiment analysis: Methods and applications | |
| Wang et al. | Chinese poetry generation with planning based neural network | |
| CN107239439A (en) | Public sentiment sentiment classification method based on word2vec | |
| CN102682130B (en) | Text sentiment classification method and system | |
| Del Bosque et al. | Aggressive text detection for cyberbullying | |
| Puniyani et al. | Social links from latent topics in microblogs | |
| CN105468713A (en) | Multi-model fused short text classification method | |
| CN103744953A (en) | Network hotspot mining method based on Chinese text emotion recognition | |
| CN103678278A (en) | Chinese text emotion recognition method | |
| CN107885883A (en) | A kind of macroeconomy field sentiment analysis method and system based on Social Media | |
| CN106202584A (en) | A kind of microblog emotional based on standard dictionary and semantic rule analyzes method | |
| Sharma et al. | An artificial neural network based approach for sentiment analysis of opinionated text | |
| CN106294316A (en) | A kind of text emotion based on dictionary analyzes method | |
| CN103995853A (en) | Multi-language emotional data processing and classifying method and system based on key sentences | |
| CN105740382A (en) | Aspect classification method for short comment texts | |
| CN112966508B (en) | Universal automatic term extraction method | |
| KR101319413B1 (en) | Summary Information Generating System and Method for Review of Product and Service | |
| CN110134934A (en) | Text emotion analysis method and device | |
| Agarwal et al. | Sentiment classification of review documents using phrase patterns | |
| CN106227768A (en) | A kind of short text opining mining method based on complementary language material | |
| CN102929860A (en) | Chinese clause emotion polarity distinguishing method based on context | |
| CN109885776A (en) | Recommended models can be explained in open source community PR reviewer | |
| CN106055633A (en) | Chinese microblog subjective and objective sentence classification method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 100044 Tengda Building, 168 Xizhimenwai Street, Haidian District, Beijing, 2101-2103 and 2105-2111 on the 21st floor Applicant after: Beijing Chehui Technology Co., Ltd. Address before: 100044 01-07, 29 story, Tengda tower, 168 west gate, Haidian District, Beijing. Applicant before: Beijing Che Hui Interactive Advertising Co., Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |