CN106649270A - Public opinion monitoring and analyzing method - Google Patents
Public opinion monitoring and analyzing method Download PDFInfo
- Publication number
- CN106649270A CN106649270A CN201611176739.3A CN201611176739A CN106649270A CN 106649270 A CN106649270 A CN 106649270A CN 201611176739 A CN201611176739 A CN 201611176739A CN 106649270 A CN106649270 A CN 106649270A
- Authority
- CN
- China
- Prior art keywords
- analyzing
- monitoring
- candidate word
- text emotion
- information
- 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
- G06F40/30—Semantic analysis
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Abstract
The invention relates to the information technology and provides a public opinion monitoring and analyzing method in order to solve the problem that existing public opinion analyzing software systems cannot specially achieve deep, detailed and quantitative evaluation of competitiveness of products and manufacturer according to public opinion information. The method includes the steps that firstly, e-commerce website commodity and comment information is captured in real time through the distributed web crawler technology, and structural data in the information is extracted through a template and stored; then according to the stored structural data, the data is automatically subjected to classification, clustering, abstract generation and name organization, and positive and negative property prejudgment is conducted; finally, the structural data is output and displayed according to requirements. The method has the advantage that reference is provided for the manufacturers or commodity developers, and is suitable for information acquisition and analysis system.
Description
Technical field
The present invention relates to information technology, more particularly to internet big data analytical technology.
Background technology
Those cannot share in big data epoch, conventional daily life, immeasurable information is all by digitization, people
Information can be obtained by internet, be participated in discussion and expression of opinion, on the one hand, for the relevant information such as brand product of enterprise
Suggestion also by internet information expressed and transmitted one after another, form network public opinion；On the other hand, in daily life
The product for touching, people also tend to carry out the inquiry of product information by internet, including the comment, specially of other users
The assessment of industry website and the advertisement of portal website etc., at the same time, user can also issue the assessment to enterprise or product.Network
The fast propagation of information and diffusion, it is possible to create huge public opinion strength.Therefore, to be in the enterprise in the big data epoch necessary
The value of data is made full use of, is excavated comprehensively and is monitored internet data information, in order to being improved to product, being innovated, more
Change and other enterprise-levels decision-making, safeguard brand image, expand brand influence, the final competitiveness for promoting enterprise.
Due to the diversity from internet mass data form, Traditional Man collection, processing data mode have been difficult to
It is competent.Although there are many analysis of public opinion software systems on the market at present, it does not all utilize public feelings information specially deeply
The careful competitiveness for product, manufacturer makes quantitative assessment.The public praise of manufacturer, product or even product attribute, competition
Power is the very valuable information being hidden in the public sentiment data of magnanimity.The target of product the analysis of public opinion is not merely to correlation
The theme of product, focus are parsed, tracked, being predicted and early warning, it is often more important that deep solution is cut in whole industry market
The relative competitive of every product of every manufacturer, and the survival of the fittest of the product for being quantified to enable whole industry is qualitative
Quantitative is clearly represented, while the good and bad point between product can also be calibrated.
The content of the invention
The invention aims to it is all specially deeply thinless using public feelings information to solve current the analysis of public opinion software systems
The competitiveness for product, manufacturer for causing makes the problem of quantitative assessment, there is provided a kind of public sentiment method for monitoring and analyzing.
The present invention solves its technical problem, and the technical scheme of employing is, public sentiment method for monitoring and analyzing, it is characterised in that bag
Include following steps：
Step 1, each electric business web site commodity and review information are captured in real time by distributed network crawler technology, using template
Extract structural data therein to be stored；
Step 2, the structural data for being stored, are classified to it, are clustered, being generated summary and title knowledge automatically
Not, and positive and negative property anticipation is carried out；
Step 3, output are simultaneously presented according to demand structural data.
Specifically, in step 2, the positive and negative property anticipation is referred to and carries out text emotion analysis to review information.
Further, it is described text emotion analysis is carried out to review information method be：
Step 201, set up different text emotion analysis models for the different type of merchandises；
Step 202, the type for judging the affiliated commodity of the review information, select the corresponding text emotion analysis of the type of merchandise
Model is analyzed.
It is specifically, described to set up in different text emotion analysis models for the different type of merchandises in step 201,
The method for building up of its text emotion analysis model is：Existing multiple review informations for a certain type of merchandise are obtained as instruction
Practice collection, Chinese word segmentation operation is carried out in the review information of training set, obtain multiple candidate words, obtain each candidate word corresponding
Sentiment orientation, using candidate word as feature text emotion analysis model is set up.
Further, the mode of the corresponding Sentiment orientation of described each candidate word of acquisition is：Judge candidate word with it is general
Semantic distance in emotion benchmark word dictionary between each emotion benchmark word, determines the Sentiment orientation of candidate word.
Specifically, the mode of the corresponding Sentiment orientation of described each candidate word of acquisition is：It is artificial to set up mark emotion language material
Storehouse, candidate word is matched with the artificial mark Emotional Corpus set up, and determines the Sentiment orientation of candidate word.
Further, in step 202, in analysis, also extract the candidate word in each review information and carry out statistics row
Sequence, deletes feature poorly efficient and/or invalid in text emotion analysis model.
Specifically, in step 2, the structural data also to being stored is cleaned, the cleaning be to Outlier Data and
Substantially irrational data are rejected.
The invention has the beneficial effects as follows, in the present invention program, by above-mentioned public sentiment method for monitoring and analyzing, can be to comment letter
Breath is analyzed automatically, is manufacturer or business so as to draw the quantitative assessment done by the competitiveness for product or manufacturer
Product developer provides reference, improving product efficiency of research and development and specific aim.
Specific embodiment
With reference to embodiment, technical scheme is described in detail.
Public sentiment method for monitoring and analyzing of the present invention is：First each electric business is captured in real time by distributed network crawler technology
Web site commodity and review information, are stored using template extraction structural data therein；Then it is directed to stored structure
Change data, it classified automatically, is clustered, generating summary and title identification, and carry out positive and negative property anticipation；Finally export simultaneously
Structural data is presented according to demand.
Embodiment
The public sentiment method for monitoring and analyzing of the embodiment of the present invention, it is comprised the following steps：
Step 1, each electric business web site commodity and review information are captured in real time by distributed network crawler technology, using template
Extract structural data therein to be stored.
In this step, distributed network crawler technology is a kind of existing more general technology for information acquisition, herein no longer
Describe in detail.
Step 2, the structural data for being stored, are classified to it, are clustered, being generated summary and title knowledge automatically
Not, and positive and negative property anticipation is carried out.
In this step, positive and negative property anticipation is referred to and carries out text emotion analysis to review information, its analysis method can be with
Lower concrete steps：
Step 201, set up different text emotion analysis models for the different type of merchandises.
Here, set up in different text emotion analysis models for the different type of merchandises, its text emotion analysis mould
The method for building up of type can be：Existing multiple review informations for a certain type of merchandise are obtained as training set, in training set
Review information in carry out Chinese word segmentation operation, obtain multiple candidate words, obtain the corresponding Sentiment orientation of each candidate word, will wait
Word is selected to set up text emotion analysis model as feature.Obtaining the mode of the corresponding Sentiment orientation of each candidate word can be：1) sentence
Semantic distance in disconnected candidate word and general emotion benchmark word dictionary between each emotion benchmark word, the emotion for determining candidate word is inclined
To；2) it is artificial to set up mark Emotional Corpus, candidate word is matched with the artificial mark Emotional Corpus set up, it is determined that waiting
Select the Sentiment orientation of word.
Step 202, the type for judging the affiliated commodity of the review information, select the corresponding text emotion analysis of the type of merchandise
Model is analyzed.
In analysis, the candidate word in each review information can also be extracted and sort method is carried out, delete text emotion point
Poorly efficient and/or invalid feature, i.e., be updated to text emotion analysis model in analysis model.
Here, in step 2, the structural data preferably also to being stored is cleaned, cleaning refer to Outlier Data and
Substantially irrational data are rejected.
Step 3, output are simultaneously presented according to demand structural data.
Here, the mode that structural data is presented according to demand is varied, is existing more ripe technology, therefore
No longer describe in detail herein.
Claims (8)
1. public sentiment method for monitoring and analyzing, it is characterised in that comprise the following steps：
Step 1, each electric business web site commodity and review information are captured in real time by distributed network crawler technology, using template extraction
Structural data therein is stored；
Step 2, the structural data for being stored, are classified to it, are clustered, being generated summary and title identification automatically, and
Carry out positive and negative property anticipation；
Step 3, output are simultaneously presented according to demand structural data.
2. public sentiment method for monitoring and analyzing as claimed in claim 1, it is characterised in that in step 2, the positive and negative property anticipation is
Finger carries out text emotion analysis to review information.
3. public sentiment method for monitoring and analyzing as claimed in claim 2, it is characterised in that described that text emotion is carried out to review information
The method of analysis is：
Step 201, set up different text emotion analysis models for the different type of merchandises；
Step 202, the type for judging the affiliated commodity of the review information, select the corresponding text emotion analysis model of the type of merchandise
It is analyzed.
4. public sentiment method for monitoring and analyzing as claimed in claim 3, it is characterised in that described for different business in step 201
Category type is set up in different text emotion analysis models, and the method for building up of its text emotion analysis model is：Obtain existing
Multiple review informations for a certain type of merchandise carry out Chinese word segmentation behaviour as training set, in the review information of training set
Make, obtain multiple candidate words, obtain the corresponding Sentiment orientation of each candidate word, using candidate word as feature text emotion point is set up
Analysis model.
5. public sentiment method for monitoring and analyzing as claimed in claim 4, it is characterised in that the corresponding feelings of each candidate word of the acquisition
Feeling the mode being inclined to is：Judge the semantic distance between each emotion benchmark word in candidate word and general emotion benchmark word dictionary, really
Determine the Sentiment orientation of candidate word.
6. public sentiment method for monitoring and analyzing as claimed in claim 4, it is characterised in that the corresponding feelings of each candidate word of the acquisition
Feeling the mode being inclined to is：It is artificial to set up mark Emotional Corpus, candidate word is carried out with the artificial mark Emotional Corpus set up
Matching, determines the Sentiment orientation of candidate word.
7. public sentiment method for monitoring and analyzing as claimed in claim 4, it is characterised in that in step 202, in analysis, also extracts
Candidate word in each review information simultaneously carries out sort method, deletes feature poorly efficient and/or invalid in text emotion analysis model.
8. the public sentiment method for monitoring and analyzing as described in claim 1 or 2 or 3 or 4 or 5 or 6 or 7, it is characterised in that in step 2,
Structural data also to being stored is cleaned, and the cleaning is that Outlier Data and obvious irrational data are picked
Remove.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611176739.3A CN106649270A (en) | 2016-12-19 | 2016-12-19 | Public opinion monitoring and analyzing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611176739.3A CN106649270A (en) | 2016-12-19 | 2016-12-19 | Public opinion monitoring and analyzing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106649270A true CN106649270A (en) | 2017-05-10 |
Family
ID=58823968
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611176739.3A Pending CN106649270A (en) | 2016-12-19 | 2016-12-19 | Public opinion monitoring and analyzing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106649270A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107293309A (en) * | 2017-05-19 | 2017-10-24 | 四川新网银行股份有限公司 | A kind of method that lifting public sentiment monitoring efficiency is analyzed based on customer anger |
| CN108647249A (en) * | 2018-04-18 | 2018-10-12 | 平安科技（深圳）有限公司 | Public sentiment data prediction technique, device, terminal and storage medium |
| CN108681977A (en) * | 2018-03-27 | 2018-10-19 | 成都律云科技有限公司 | A kind of lawyer's information processing method and system |
| CN108874992A (en) * | 2018-06-12 | 2018-11-23 | 深圳华讯网络科技有限公司 | The analysis of public opinion method, system, computer equipment and storage medium |
| CN109376237A (en) * | 2018-09-04 | 2019-02-22 | 中国平安人寿保险股份有限公司 | Prediction technique, device, computer equipment and the storage medium of client's stability |
| CN109522466A (en) * | 2018-10-20 | 2019-03-26 | 河南工程学院 | A kind of distributed reptile system |
| CN115374332A (en) * | 2022-09-06 | 2022-11-22 | 北京化工大学 | Emergency rescue resource retrieval method, device and equipment |
| CN116188103A (en) * | 2023-02-07 | 2023-05-30 | 杭州展俊科技有限公司 | Big data intelligent replenishment processing method for cross-border electronic commerce |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103207855A (en) * | 2013-04-12 | 2013-07-17 | 广东工业大学 | Fine-grained sentiment analysis system and method specific to product comment information |
| CN103365867A (en) * | 2012-03-29 | 2013-10-23 | 腾讯科技（深圳）有限公司 | Method and device for emotion analysis of user evaluation |
| CN106127507A (en) * | 2016-06-13 | 2016-11-16 | 四川长虹电器股份有限公司 | A kind of commodity the analysis of public opinion method and system based on user's evaluation information |
-
2016
- 2016-12-19 CN CN201611176739.3A patent/CN106649270A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103365867A (en) * | 2012-03-29 | 2013-10-23 | 腾讯科技（深圳）有限公司 | Method and device for emotion analysis of user evaluation |
| CN103207855A (en) * | 2013-04-12 | 2013-07-17 | 广东工业大学 | Fine-grained sentiment analysis system and method specific to product comment information |
| CN106127507A (en) * | 2016-06-13 | 2016-11-16 | 四川长虹电器股份有限公司 | A kind of commodity the analysis of public opinion method and system based on user's evaluation information |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107293309A (en) * | 2017-05-19 | 2017-10-24 | 四川新网银行股份有限公司 | A kind of method that lifting public sentiment monitoring efficiency is analyzed based on customer anger |
| CN108681977A (en) * | 2018-03-27 | 2018-10-19 | 成都律云科技有限公司 | A kind of lawyer's information processing method and system |
| CN108681977B (en) * | 2018-03-27 | 2022-05-31 | 成都律云科技有限公司 | Lawyer information processing method and system |
| CN108647249A (en) * | 2018-04-18 | 2018-10-12 | 平安科技（深圳）有限公司 | Public sentiment data prediction technique, device, terminal and storage medium |
| CN108874992A (en) * | 2018-06-12 | 2018-11-23 | 深圳华讯网络科技有限公司 | The analysis of public opinion method, system, computer equipment and storage medium |
| CN108874992B (en) * | 2018-06-12 | 2021-03-19 | 深圳华讯网络科技有限公司 | Public opinion analysis method, system, computer equipment and storage medium |
| CN109376237A (en) * | 2018-09-04 | 2019-02-22 | 中国平安人寿保险股份有限公司 | Prediction technique, device, computer equipment and the storage medium of client's stability |
| CN109522466A (en) * | 2018-10-20 | 2019-03-26 | 河南工程学院 | A kind of distributed reptile system |
| CN115374332A (en) * | 2022-09-06 | 2022-11-22 | 北京化工大学 | Emergency rescue resource retrieval method, device and equipment |
| CN116188103A (en) * | 2023-02-07 | 2023-05-30 | 杭州展俊科技有限公司 | Big data intelligent replenishment processing method for cross-border electronic commerce |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106649270A (en) | Public opinion monitoring and analyzing method | |
| Gokulakrishnan et al. | Opinion mining and sentiment analysis on a twitter data stream | |
| CN104572958B (en) | A kind of sensitive information monitoring method based on event extraction | |
| Venugopalan et al. | Exploring sentiment analysis on twitter data | |
| CN104598535B (en) | A kind of event extraction method based on maximum entropy | |
| TWI424325B (en) | Systems and methods for organizing collective social intelligence information using an organic object data model | |
| CN109829166B (en) | People and host customer opinion mining method based on character-level convolutional neural network | |
| CN105550269A (en) | Product comment analyzing method and system with learning supervising function | |
| CN103064971A (en) | Scoring and Chinese sentiment analysis based review spam detection method | |
| Yussupova et al. | Models and methods for quality management based on artificial intelligence applications | |
| Halibas et al. | Application of text classification and clustering of Twitter data for business analytics | |
| CN103646088A (en) | Product comment fine-grained emotional element extraction method based on CRFs and SVM | |
| CN106354845A (en) | Microblog rumor recognizing method and system based on propagation structures | |
| JP2011204226A (en) | System and method for classifying text feeling polarities based on sentence sequence | |
| CN103853824A (en) | In-text advertisement releasing method and system based on deep semantic mining | |
| CN106407236A (en) | An emotion tendency detection method for comment data | |
| CN112069312B (en) | Text classification method based on entity recognition and electronic device | |
| CN110321549B (en) | New concept mining method based on sequential learning, relation mining and time sequence analysis | |
| CN112163424A (en) | Data labeling method, device, equipment and medium | |
| CN107832781A (en) | A kind of software defect towards multi-source data represents learning method | |
| CN106407235A (en) | A semantic dictionary establishing method based on comment data | |
| CN110910175A (en) | Tourist ticket product portrait generation method | |
| Meng et al. | Mining user reviews: from specification to summarization | |
| CN106897274B (en) | Cross-language comment replying method | |
| CN112115712B (en) | Topic-based group emotion analysis 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170510 |
|
| RJ01 | Rejection of invention patent application after publication |