CN107748743A - A kind of electric business online comment text emotion analysis method - Google Patents
A kind of electric business online comment text emotion analysis method Download PDFInfo
- Publication number
- CN107748743A CN107748743A CN201710849722.8A CN201710849722A CN107748743A CN 107748743 A CN107748743 A CN 107748743A CN 201710849722 A CN201710849722 A CN 201710849722A CN 107748743 A CN107748743 A CN 107748743A
- Authority
- CN
- China
- Prior art keywords
- emotion
- sentence
- online comment
- electric business
- value
- 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
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- 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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3335—Syntactic pre-processing, e.g. stopword elimination, stemming
-
- 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
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a kind of electric business online comment text emotion analysis method, belong to natural language processing technique and Judgment by emotion field, build electric business online comment sentiment dictionary, realize electric business online comment Text Pretreatment into sentence unit first, sentence unit is handled using 2 metagrammars in n gram patterns again, the emotion value p after processingiAddition obtains the emotion basic value p of sentence unit after collecting, generate p list, solves p synthesis, carries out Judgment by emotion, helps enterprise more rationally to make marketing strategy.
Description
Technical field
The invention belongs to natural language processing technique and Judgment by emotion field, and in particular to a kind of electric business online comment text
Sentiment analysis method.
Background technology
With the sound development of Electronic Commerce in China, flow bonus phase mistake, client cost more and more higher.Ecommerce is looked forward to
How industry identifies client's consumption preferences, carries out precision marketing, reduces cost of competition, is each enterprise's indispensability homework.Online comment
Data are evaluation of the client to the quality of product, price, service etc. after the completion of e-commerce transaction.Online comment data
Turn into enterprise to obtain client's consumption preferences, carry out the important information source of precision marketing.This evaluate collection is often with very strong
Sentiment orientation.The Sentiment orientation of research client can measure degree of recognition of the client to enterprise, and the consumption that can also excavate client is inclined
It is good.
A kind of accurate electric business online comment sentiment analysis method of science, can allow enterprise quickly grasp client consumption it is inclined
It is good, the deficiency of oneself is found, improves enterprise competitiveness.However, in existing technology, it is accurate not solve text emotion analysis
The problem of rate is low, emotion changes with monitoring in real time, it is also difficult to the value of one client of accurate evaluation.
Therefore, a kind of electric business text online comment sentiment analysis method, is current urgent problem.
The content of the invention
According to above the deficiencies in the prior art, the present invention proposes a kind of electric business online comment text emotion analysis method, leads to
2 metagrammars crossed in n-gram are handled electric business online comment sentiment dictionary, have effectively grasped the hobby of client, favorably
In equilibrium of the product in market.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention is：A kind of electric business online comment text emotion
Analysis method, it is characterised in that comprise the following steps：
Step 1：Electric business online comment sentiment dictionary is built, sentiment dictionary includes positive emotion set of words, negative emotion
Set of words, negative set of words and degree set of words；
Step 2：Each customer's online comment text in electric business online comment is pre-processed, to text segmentation subordinate sentence
Sentence is obtained, sentence is divided into by sentence unit by punctuation mark, then the word in sentence unit segmented, is formed
Phrase；
Step 3：N-gram processing is carried out to the phrase obtained in step 2, using 2 metagrammars in n-gram to sentence list
Phrase in member is combined to form 2 yuan of phrases, and the emotion basic value p of sentence unit is calculated with reference to electric business online comment sentiment dictionary；
Step 4：Each customer's online comment is split in electric business online comment is split as sentence list again for sentence, sentence
Member, therefore each customer's online comment, according to the emotion basic value p of each sentence unit, one emotion value p list of generation is p-
list；
Step 5：According to the p-list of each customer's online comment, carry out addition and collect to calculate p- synthesis；
Step 6：Results contrast is carried out with 0 using the p- synthesis in step 5, the online comment to each customer carries out emotion
Tendency judges.
In above method step, the detailed process of the step 3 is：Step 3.1：2 metagrammars are to the word in sentence unit
Group is combined, and forms 2 yuan of phrases, and that sentences anterior locations is referred to as anteposition word, and that sentences back location is referred to as rear position word；Step
3.2：The calculating of positive emotion set of words and negative emotion set of words progress emotion value p is sentenced in position word combination sentiment dictionary afterwards
It is disconnected；Anteposition word combine negative set of words judge emotion value p whether * (- 1), anteposition word combination degree set of words judges that emotion value p is
No * (- 1)；As a result an emotion value p is obtained for each 2 yuan of phrasesi；Step 3.3：By 2 yuan of phrases in whole sentence unit
Emotion value piAddition obtains the basic value p of sentence unit after collecting.
In above method step, the p- synthesis in the step 5 is that the statistics read group total of p value in p-list obtains.
Present invention has the advantages that：It is low that the present invention solves text emotion analysis accuracy rate in electric business online comment analysis
The problem of, the value of client can be assessed, can be that the change of follow-up emotion is carried out with monitoring, Customer Satisfaction differentiation and precision marketing
Place mat；Using the present invention, enterprise can be allowed quickly to understand the Sentiment orientation of online comment, improve purchase experiences, accurately make battalion
Sell decision-making；Also problem present in service of goods can be identified rapidly, is improved service quality, and strengthen enterprise competitiveness.
Brief description of the drawings
The content expressed by this specification accompanying drawing and the mark in figure are briefly described below：
Fig. 1 is the flow chart of the embodiment of the present invention.
Embodiment
Below by the description to embodiment, the shape of for example involved each component of embodiment of the invention, structure
Make, the mutual alignment between each several part and annexation, the effect of each several part and operation principle, manufacturing process and the side of operating with
Method etc., is described in further detail, completeer to help those skilled in the art to have inventive concept of the invention, technical scheme
Whole, accurate and deep understanding.
A kind of electric business online comment text emotion analysis method, comprises the following steps：
Step 1：Electric business online comment sentiment dictionary is built, sentiment dictionary includes positive emotion set of words, negative emotion
Set of words, negative set of words and degree set of words；By taking positive emotion set of words as an example, form is following (partial words)：' no
It is wrong ', ' happy ', ' satisfaction ', ' joyful ', the word such as ' stick '.The set of emotion word has the distribution side of acquiescence in this profession
Formula, it is common database, here, repeating no more.
Step 2：Electric business online comment Text Pretreatment, including sentence is obtained to text segmentation subordinate sentence, pass through punctuation mark
(', ', '！’、‘.’、‘；', ', ' etc. punctuation mark) sentence unit that is divided into sentence, then to the word in sentence unit
Segmented, form phrase.
By taking txt texts as an example, form is following (being the online comment case of a customer below)：Text original text：" send part too
It is fast, it is stupefied by such speed for the first time, it is super happy！”
Cutting is as follows for phrase form after sentence unit：
The phrase of sentence unit 1：Send part too fast
The phrase of sentence unit 2：It is stupefied by such speed for the first time
The phrase of sentence unit 3：It is super happy
Step 3：N-gram processing is carried out to phrase, the phrase in sentence unit entered using 2 metagrammars in n-gram
2 yuan of phrases are formed after row combination, calculate the emotion basic value p of sentence unit.
Particular content is：Step 3.1：2 metagrammars are combined to the phrase in sentence unit, after combination, sentence before
Position is referred to as anteposition word, and that sentences back location is referred to as rear position word.Such as 2 metagrammars that " text " and " emotion " combines
" text emotion ", wherein " text " is anteposition word, position word after " emotion " is referred to as.
Step 3.2：Positive emotion set of words and negative emotion set of words enter market in position word combination sentiment dictionary afterwards
The calculating for feeling basic value p judges that p initial values are set as 0, and emotion basic value is p+1 when positive emotion set of words occurs,
Emotion basic value p-1 when negative emotion set of words occurs, therefore p can be 1,0 or -1.Anteposition word combines negative set of words and entered
Row, which calculates, to be judged, if there is the word in negative word dictionary, then emotion basic value p* (- 1)；If occurred without, emotion base
This value p is constant；Judge whether occur degree word in anteposition word while the emotion basic value of anteposition word calculates or after calculating
Word (anteposition word emotion basic value and the calculating section grain husk of judgment value do not consider priority) in set, if there is degree word word
Word in allusion quotation, then anteposition word combination degree set of words continue to calculate p value, p* (degree word assign weight)；If do not go out
Existing degree word, then p value is constant.
Judgement of the word in set, according to the principle of screening, it is first determined whether positive emotion set of words is appeared in,
It is positive emotion word if occurring in positive emotion set of words, if do not occurred；Judge to comment successively according to the method
Negative emotion set of words, negative set of words and degree set of words whether are appeared in by word, if not in these set of words
In, then do not make a decision.After the determined property for commenting on word, and then calculate emotion basic value p.
Step 3.3：Emotion value p corresponding to the 2 yuan of phrases that will be formed in whole sentence unit based on 2 metagrammarsiIt is added
The p value of sentence unit is obtained after collecting.Because sentence has been cut into very small unit, often only go out in sentence unit
Existing single emotional, so all emotion value p calculated by 2 metagrammars in a sentence unitiAddition obtains the sentence after collecting
The emotion basic value p of subelement.
The result of above-described embodiment sentence unit 1 is as follows：
2 yuan of phrases in sentence unit 1：Send part too fast too fast
The p value of sentence unit 1：1.Wherein, p1Represent the emotion basic calculating value of " send part too fast ", p2Represent " too fast "
Emotion basic calculating value.p1Initial value is 0, and " send part too fast " is used as a 2 metagrammar structures, and " too fast " is in positive emotion word
Occur in language set, p1=0+1=1, " sending part " does not occur in negative set of words, degree set of words, therefore p1Continue as 1；p2Just
Initial value is 0, and " too fast " to be used as another 2 metagrammar structure, " " does not occur in positive emotion set of words, p2Still
For 0；The p value of sentence unit 1 is p1+p2=1+0=1.Wherein p1The emotion value 1, p corresponding to " send part too fast "2For " too fast "
Corresponding emotion value 0.
Step 4：Each customer's online comment is split in electric business online comment is split as sentence list again for sentence, sentence
Member, therefore each customer's online comment is actually made up of multiple sentence units.Therefore each customer's online comment is according to each sentence
The emotion basic value p of subelement, one emotion value p list of generation is p-list.
Each customer's online comment is split in electric business online comment utilizes 2 for many sentence units, each sentence unit
Metagrammar and sentiment dictionary, an emotion basic value p is drawn, therefore each customer's online comment is formed by many emotion basic value p
One list, i.e. p-list.The p-list of sentence unit 1-3 embodiments in above-mentioned steps 2 is [1,1,2].
Step 5：P- synthesis is calculated according to the list p-list of text emotion value.For the p in step 4 in p-list
Value, carries out statistics read group total, and summed result integrates for p-.The p- overall targets of embodiment are 4 in above-mentioned steps 2.
Step 6：Results contrast is carried out with 0 using the p- synthesis in step 5, to the carry out emotion of each customer's online comment
Tendency judges to carry out Sentiment orientation judgement.According to p- synthesis and 0 comparative result progress Sentiment orientation judgement, i.e. p- synthesis>0 is
Actively, p- is integrated<0 is passiveness, and p- synthesis=0 is neutrality.Customer's online comment Judgment by emotion result in step 2 in embodiment
To be positive.
The present invention is exemplarily described above, it is clear that present invention specific implementation is not subject to the restrictions described above,
As long as employing the improvement of the various unsubstantialities of inventive concept and technical scheme of the present invention progress, or not improved this is sent out
Bright design and technical scheme directly applies to other occasions, within protection scope of the present invention.The protection of the present invention
Scope should be determined by the scope of protection defined in the claims.
Claims (3)
1. a kind of electric business online comment text emotion analysis method, it is characterised in that comprise the following steps：
Step 1：Electric business online comment sentiment dictionary is built, sentiment dictionary includes positive emotion set of words, negative emotion word
Set, negative set of words and degree set of words；
Step 2：Each customer's online comment text in electric business online comment is pre-processed, text segmentation subordinate sentence is obtained
Sentence, sentence is divided into by sentence unit by punctuation mark, then the word in sentence unit segmented, form word
Group；
Step 3：N-gram processing is carried out to the phrase obtained in step 2, using 2 metagrammars in n-gram in sentence unit
Phrase combine to form 2 yuan of phrases, with reference to electric business online comment sentiment dictionary calculate sentence unit emotion basic value p；
Step 4：Each customer's online comment is split as sentence in electric business online comment, and sentence is split as sentence unit again,
Therefore each customer's online comment, according to the emotion basic value p of each sentence unit, one emotion value p list of generation is p-list；
Step 5：According to the p-list of each customer's online comment, carry out addition and collect to calculate p- synthesis；
Step 6：Results contrast is carried out with 0 using the p- synthesis in step 5, the online comment to each customer carries out Sentiment orientation
Judge.
2. electric business online comment text emotion analysis method according to claim 1, it is characterised in that the tool of the step 3
Body process is：
Step 3.1：2 metagrammars are combined to the phrase in sentence unit, form 2 yuan of phrases, sentence being referred to as anterior locations
Anteposition word, that sentences back location is referred to as rear position word；
Step 3.2：Positive emotion set of words and negative emotion set of words carry out emotion value p in position word combination sentiment dictionary afterwards
Calculating judge；Anteposition word combine negative set of words judge emotion value p whether * (- 1), anteposition word combination degree set of words judgement
Emotion value p whether * (- 1)；As a result an emotion value p is obtained for each 2 yuan of phrasesi；
Step 3.3：By the emotion value p of 2 yuan of phrases in whole sentence unitiAddition obtains the basic value p of sentence unit after collecting.
3. electric business online comment text emotion analysis method according to claim 1, it is characterised in that in the step 5
P- synthesis is that the statistics read group total of p value in p-list obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710849722.8A CN107748743A (en) | 2017-09-20 | 2017-09-20 | A kind of electric business online comment text emotion analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710849722.8A CN107748743A (en) | 2017-09-20 | 2017-09-20 | A kind of electric business online comment text emotion analysis method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107748743A true CN107748743A (en) | 2018-03-02 |
Family
ID=61255348
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710849722.8A Pending CN107748743A (en) | 2017-09-20 | 2017-09-20 | A kind of electric business online comment text emotion analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107748743A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108376134A (en) * | 2018-04-19 | 2018-08-07 | 安徽商贸职业技术学院 | A kind of electric business online comment text word analysis method based on position-order statistics |
| CN108596637A (en) * | 2018-04-24 | 2018-09-28 | 北京航空航天大学 | A kind of electric business service problem discovery system |
| CN109816443A (en) * | 2019-01-18 | 2019-05-28 | 安徽商贸职业技术学院 | A kind of user's pain spot quantization method based on sentiment analysis |
| CN109857852A (en) * | 2019-01-24 | 2019-06-07 | 安徽商贸职业技术学院 | A kind of the screening judgment method and system of electric business online comment training set feature |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663046A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院自动化研究所 | Sentiment analysis method oriented to micro-blog short text |
| CN103123633A (en) * | 2011-11-21 | 2013-05-29 | 阿里巴巴集团控股有限公司 | Generation method of evaluation parameters and information searching method based on evaluation parameters |
| CN103150367A (en) * | 2013-03-07 | 2013-06-12 | 宁波成电泰克电子信息技术发展有限公司 | Method for analyzing emotional tendency of Chinese microblogs |
| CN103559233A (en) * | 2012-10-29 | 2014-02-05 | 中国人民解放军国防科学技术大学 | Extraction method for network new words in microblogs and microblog emotion analysis method and system |
| CN105095190A (en) * | 2015-08-25 | 2015-11-25 | 众联数据技术（南京）有限公司 | Chinese semantic structure and finely segmented word bank combination based emotional analysis method |
| WO2017042620A1 (en) * | 2015-09-08 | 2017-03-16 | Iacus Stefano Maria | Isa: a fast, scalable and accurate algorithm for supervised opinion analysis |
-
2017
- 2017-09-20 CN CN201710849722.8A patent/CN107748743A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103123633A (en) * | 2011-11-21 | 2013-05-29 | 阿里巴巴集团控股有限公司 | Generation method of evaluation parameters and information searching method based on evaluation parameters |
| CN102663046A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院自动化研究所 | Sentiment analysis method oriented to micro-blog short text |
| CN103559233A (en) * | 2012-10-29 | 2014-02-05 | 中国人民解放军国防科学技术大学 | Extraction method for network new words in microblogs and microblog emotion analysis method and system |
| CN103150367A (en) * | 2013-03-07 | 2013-06-12 | 宁波成电泰克电子信息技术发展有限公司 | Method for analyzing emotional tendency of Chinese microblogs |
| CN105095190A (en) * | 2015-08-25 | 2015-11-25 | 众联数据技术（南京）有限公司 | Chinese semantic structure and finely segmented word bank combination based emotional analysis method |
| WO2017042620A1 (en) * | 2015-09-08 | 2017-03-16 | Iacus Stefano Maria | Isa: a fast, scalable and accurate algorithm for supervised opinion analysis |
Non-Patent Citations (2)
| Title |
|---|
| 任巨伟 等: "二元情感常识库建设及其在文本情感分析", 《中国科技论文在线》 * |
| 杜振雷: "面向微博短文本的情感分析研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108376134A (en) * | 2018-04-19 | 2018-08-07 | 安徽商贸职业技术学院 | A kind of electric business online comment text word analysis method based on position-order statistics |
| CN108596637A (en) * | 2018-04-24 | 2018-09-28 | 北京航空航天大学 | A kind of electric business service problem discovery system |
| CN108596637B (en) * | 2018-04-24 | 2022-05-06 | 北京航空航天大学 | Automatic E-commerce service problem discovery system |
| CN109816443A (en) * | 2019-01-18 | 2019-05-28 | 安徽商贸职业技术学院 | A kind of user's pain spot quantization method based on sentiment analysis |
| CN109857852A (en) * | 2019-01-24 | 2019-06-07 | 安徽商贸职业技术学院 | A kind of the screening judgment method and system of electric business online comment training set feature |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109165294B (en) | Short text classification method based on Bayesian classification | |
| CN107748743A (en) | A kind of electric business online comment text emotion analysis method | |
| WO2022141861A1 (en) | Emotion classification method and apparatus, electronic device, and storage medium | |
| CN103365867B (en) | It is a kind of that the method and apparatus for carrying out sentiment analysis are evaluated to user | |
| CN111666761B (en) | Fine-grained emotion analysis model training method and device | |
| Oliveira et al. | Some experiments on modeling stock market behavior using investor sentiment analysis and posting volume from Twitter | |
| CN102929860B (en) | Chinese clause emotion polarity distinguishing method based on context | |
| CN109190121A (en) | Car review sentiment analysis method based on automobile body and part-of-speech rule | |
| CN109508373A (en) | Calculation method, equipment and the computer readable storage medium of enterprise's public opinion index | |
| CN102541838B (en) | Method and equipment for optimizing emotional classifier | |
| CN106202481A (en) | The evaluation methodology of a kind of perception data and system | |
| CN107844558A (en) | The determination method and relevant apparatus of a kind of classification information | |
| CN113360647B (en) | 5G mobile service complaint source-tracing analysis method based on clustering | |
| CN107767273B (en) | Asset configuration method based on social data, electronic device and medium | |
| CN112015721A (en) | E-commerce platform storage database optimization method based on big data | |
| JP4600045B2 (en) | Opinion extraction learning device and opinion extraction classification device | |
| CN110674296B (en) | Information abstract extraction method and system based on key words | |
| CN108090211A (en) | Hot news method for pushing and device | |
| WO2020238363A1 (en) | Method and apparatus for pushing information, electronic commerce system and storage medium | |
| CN109948148A (en) | A kind of text information emotion determination method and decision maker | |
| CN109858034A (en) | A kind of text sentiment classification method based on attention model and sentiment dictionary | |
| CN110543547A (en) | automobile public praise semantic emotion analysis system | |
| CN106055633A (en) | Chinese microblog subjective and objective sentence classification method | |
| CN108563647A (en) | A kind of automobile Method for Sales Forecast method based on comment sentiment analysis | |
| CN115293794A (en) | Software cost evaluation method and system based on intelligent scale recognition |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180302 |