CN109145302A - Large agricultural product investor fear mood Measurement Method based on semantic text - Google Patents
Large agricultural product investor fear mood Measurement Method based on semantic text Download PDFInfo
- Publication number
- CN109145302A CN109145302A CN201811009600.9A CN201811009600A CN109145302A CN 109145302 A CN109145302 A CN 109145302A CN 201811009600 A CN201811009600 A CN 201811009600A CN 109145302 A CN109145302 A CN 109145302A
- Authority
- CN
- China
- Prior art keywords
- mood
- text
- keyword
- value
- investor
- 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.)
- Withdrawn
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
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- 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
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Abstract
The invention discloses a kind of large agricultural product investor fear mood Measurement Method based on semantic text.This method is acquired and is analyzed by some text informations relevant to large market for farm products issued to investor on internet, the text information of acquisition is pre-processed, extract mood keyword go forward side by side market thread value calculating, to obtain entire text mood value, and then propose a kind of text mood tendency decision rule, it carries out text mood tendency to determine, to realize the purpose that investor's fear mood is estimated.
Description
Technical field
The present invention relates to network technique field more particularly to a kind of large agricultural product investor based on semantic text are panic
Mood Measurement Method.
Background technique
In recent years, the specific gravity that large agricultural product account in commodity farming is increasing, the throwing of large market for farm products
Money trade development is become better and better.Large agricultural product refer to occupies greater weight in commodity farming structure, and output disappears
The biggish agricultural product such as Fei Liang, volume of trade, freight volume.Such as oilseeds, rapeseed meal, rapeseed dregs, vegetable seed, rapeseed, rape oil, rapeseed oil,
Cotton dregs, Cottonseed Meal, cottonseed, cotton oil, cottonseed oil, soya-bean oil, dregs of beans, palm oil, peanut etc..Compared with industrial products, large agricultural production
Product are mainly characterized by using soil as basic production basis, and cultivated area is very big, and planting season is longer, is easy to be existed by including weather
Interior various disasters influence, and injured area is quite big if disaster occurs, while losing also larger and being difficult to for the year
It makes up.For the investor of large agricultural product, it may be said that one, which is alarmed by the rustling of leaves, will affect and shake its investment confidence and resolution.
With popularizing for mobile Internet, social network plays an increasingly important role in people's lives,
More and more investor's selections take market information by network to harvest, and by internet platform, investor can send out at any time
Table exchanges investment experiences to the view in market and with other people.Under this trend, the network media is increasingly becoming investor's progress
The information source of investment activity and the important references for making investment decision, investor issue oneself to agricultural production on internet platform
The view in product market is customary, it may be said that the emergence and development of internet to investor participate in the investment of large agricultural product and
Decision process generates very important influence, however sometimes passively information receive also can psychology to investor and mood produce
It is raw to influence.
The mood (such as actively optimistic, passive panic mood) of investor exactly reflects expection of the investor to market,
To be had an impact to market.The mood of investor is different, influences to market bring also different.Especially passive panic feelings
Thread can move towards the market investment of large agricultural product to cause immeasurable influence.The mood and viewpoint of investor often can be with
It is come out expressed by the text issued on the network media by it.Investor is flat similar to forum, community and various blogs
Issued a large amount of valuable information on platform, and these information have the characteristics that one it is common --- be all non-structured text
Information.It may solely be invested from each expression comprising much having the mood and viewpoint sentence of expression investor in these texts
The sentence of person's mood mood, it is possible to determine that the mood of sentence investor, but to determine the investor sentiment of entire text, it is single
Determine it is inaccurate from the mood of each sentence.
For above, the present invention proposes a large agricultural product fear mood Measurement Method based on semantic text, passes through
The some text informations relevant to large market for farm products issued to investor on internet are acquired and analyze, to acquisition
Text information pre-processed, extract mood keyword and go forward side by side the calculating of market thread value, to obtain entire text mood value, into
And propose a kind of text mood tendency rule, carrying out text mood tendency determines, to realize what investor's fear mood was estimated
Purpose.
Summary of the invention
Large agricultural product investor fear feelings based on semantic text that the technical problem to be solved in the present invention is to provide a kind of
Thread Measurement Method.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention is that, large agricultural product based on semantic text
Investor's fear mood Measurement Method, comprising the following steps:
(1) investment text information acquisition: text information library is invested by establishing, using the form of web crawlers from internet
Related text content is acquired in each large agricultural product investment forum, blog;
(2) text information pre-processes: after investment text information library is established, need to pre-process text information, it is pre- to locate
Reason includes punctuate, participle, and mood keyword extraction is prepared for subsequent semantic mood analysis;
(3) semantic mood analysis: the calculating of mood value is carried out to the mood keyword that step (2) is extracted, and then is obtained whole
The mood value of a text;
Mood value is made of emotional valence and emotional intensity:
Emotional valence be divided by mood tendency it is positive and negative, wherein positive positive expectations and the emotional valence of neutral mood are "+",
The emotional valence of passive fear mood is "-";
Emotional intensity is by specific mood keyword according to the assignment rule value of table 1:
Table 1: the assignment rule of mood keyword
The mood value of each mood keyword in text is obtained by the assignment rule of table 1, and then institute in text is in love
The mood value of thread keyword carries out aggregation summation and calculates mean value, obtains the mood value of entire text, calculation formula is as follows:
Wherein E (T) indicates the mood value of a text chapter, SiIndicate i-th of mood keyword in text, E (Si) indicate
The mood value of i-th of mood keyword, n indicate the quantity of mood keyword in text;
(4) text mood tendency determines:
Table 2: text mood is inclined to decision rule
Text mood value | Text mood tendency |
Just | It is actively optimistic or neutral |
It is negative | It is passive panic |
It is compared, is determined whole by the obtained text mood value of step (3) and the text mood tendency decision rule of table 2
Whether a text is passive panic mood.
Preferably, in step (1), the acquisition of investment text information specifically includes the following steps:
(5) target topic and target acquisition web page interlinkage are determined;According to large agricultural product investor investment of user's input
Keyword is commented on as target topic, automatic search obtains relevant forum, blog network address as target and acquires website;In target
In multiple web page interlinkages that acquisition website includes, the corresponding target webpage link of target topic is determined;
(6) processing is filtered to determining target acquisition web page interlinkage;It, can in determining target acquisition web page interlinkage
It can include repetition, invalid web page interlinkage, need to be filtered processing;
(7) web page contents downloading is carried out to the target webpage link after filtration treatment, is marked according to the HTML in web page contents
Label, positioning need the corresponding URL of article acquired, the corresponding URL of the article acquired as needed, text envelope corresponding to article
Breath is downloaded and is saved in investment text information library.
Preferably, in step (2), text information pretreatment specifically includes the following steps:
(8) punctuate is handled: using Chinese fullstop or other punctuation marks as punctuate node, being needed before carrying out semantic mood analysis
Punctuate processing is carried out to collected text information;
(9) participle and part-of-speech tagging: the text after punctuate is segmented, using Python3.0 during participle
Software is segmented, the foundation of participle is Chinese corpus dictionary, using the HanLP natural language processing packet of open source, in participle
It is automatically performed part-of-speech tagging in the process, the result of participle is each phrase;
(10) it extracts mood keyword and constructs mood dictionary: for convenience of mood keyword extraction, needing to construct mood dictionary,
By all document mood phrases obtained to participle in step (9) and part-of-speech tagging, these phrases are counted, are obtained in network number
Occurs the highest phrase of word frequency in as mood keyword;Mood dictionary is added in the mood keyword of extraction, and subsequent
Constantly expand.Phrase after segmenting in text is compared with the phrase in mood dictionary, comparing successfully is that mood is closed
Keyword extracts mood keyword.
Mood dictionary creation is using CBOW in word2vector model, including input layer, projection layer, output layer;It is wherein defeated
Entering layer is 2c term vector in context (w), V (Context (w)1)、V(Context(w)2)...V(Context(w)2c);
Projection layer is then the sum of adding up for 2c term vector of input layer, and the corresponding binary tree of output layer, it is to occur in text
For the word crossed as leaf node, the number occurred in the text with each word works as the Huffman tree that weight constructs;Input layer,
Matrix-vector operation method is used between projection layer and output layer, the leaf node in output layer can all generate one due to branch
Probability, these probability multiplications can be obtained by the probability of related phrase, wherein the meter of the probability by one phrase of context-prediction
Calculation method is as follows:
P(wi| Context)=P (wi|wi-k, wi-k+1..., wi-1, wi+1..., wi+k)
In above formula, P is phrase probability, wiIndicate some word in text.
The beneficial effects of the present invention are:
It, can be to mutual by the invention of such large agricultural product investor fear mood Measurement Method based on semantic text
The mood value that large agricultural product investment text information effectively and rapidly acquire and can carry out text in networking calculates and disappears
Extremely panic mood determines, facilitates investment market to be adjusted in advance, avoids panic mood to agricultural product investment market bring
Adverse effect, the perfect prediction to large agricultural product investment market promote the development of investment market healthy and stable.
Detailed description of the invention
The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
Fig. 1 is the general flow chart of the embodiment of the present invention.
Fig. 2 is the mood dictionary construction CBOW model structure of the embodiment of the present invention.
Specific embodiment
Shown in Fig. 1, a kind of large agricultural product investor fear mood Measurement Method based on semantic text, including investment text
This information collection, text information pretreatment, semantic mood analysis, text mood tendency determine four steps:
Step 1: investment text information acquisition initially sets up investment text information library, secondly some reliable from internet
(such as major agricultural product investment forum, blog) acquisition relevant textual information in higher large agricultural product investment comment document is spent,
The text information of acquisition is finally put into investment text information library.Detailed process is as follows for acquisition:
Step 1a: target topic and target acquisition web page interlinkage are determined.It is thrown first according to large agricultural product of user's input
Money person invests comment keyword as target topic;Then automatic search obtains relevant forum, Blog Website is adopted as target
Collect website;Finally using target acquisition website as target, filter out it is relevant to target topic it is a series of link, as target acquire net
Page link.Target acquisition web page interlinkage can have one or more, the content and mesh that each target acquisition web page interlinkage includes
It is related to mark theme.
Step 1b: processing is filtered to determining target acquisition web page interlinkage.Determining the corresponding target of target topic
After web page interlinkage, the correctness of target webpage link is analyzed, correct web page interlinkage is selected, deletes duplicate webpage
Link, invalid web pages link, to improve the efficiency of acquisition.
Step 1c: web page contents downloading is carried out to the target webpage link after filtration treatment, according in web page contents
Html tag, positioning need the corresponding URL of article that acquires, the article acquired as needed corresponding URL, corresponding to article
Content of text is downloaded and is saved in investment text information library.
Step 2: text information pretreatment.Due to the information more redundancy that step 1 acquires, cannot be used directly to analyze,
It needs to pre-process the information in text library, specific step is as follows for entire preprocessing process:
Step 2a: punctuate processing.Firstly, text is imported program, fine granularity analysis, journey are then carried out as unit of sentence
Sequence can carry out punctuate processing to document automatically by the Chinese punctuation mark occurred in shelves and space etc., encounter punctuation mark fullstop
When, it is just identified as in short.
Step 2b: participle and part-of-speech tagging.Document through punctuate processing, program can call ICTCLAS3.0 segmenter automatically
Document is segmented, segmenter traverses each word in document first, and routine call Chinese corpus dictionary divides sentence
Word is simultaneously automatically performed part-of-speech tagging;After participle terminates, there is mistakes and omissions phenomenon in word segmentation result in order to prevent, finally carries out manually multiple
It looks into, the word of mistakes and omissions is found out, is segmented and is added in word segmentation result manually, word segmentation result is each phrase.
Step 2c: it extracts mood keyword and constructs mood dictionary: for convenience of mood keyword extraction, needing to construct mood word
Allusion quotation counts these phrases by all document mood phrases obtained to participle in step 2b and part-of-speech tagging, obtains in network
Occurs the highest phrase of word frequency in data as mood keyword;By the mood keyword addition mood dictionary of extraction, and after
It is continuous constantly to expand.Phrase after segmenting in text is compared with the phrase in mood dictionary, comparing successfully is mood
Keyword extracts mood keyword;
The mood dictionary creation is using CBOW in word2vector model, including input layer, projection layer, output layer;Its
Middle input layer is 2c term vector in context (w), V (Context (w)1)、V(Context(w)2)...V(Context
(w)2c);Projection layer is then the sum of adding up for 2c term vector of input layer, and the corresponding binary tree of output layer, it is with text
For the middle word occurred as leaf node, the number occurred in the text with each word works as the Huffman tree that weight constructs;It is defeated
Enter and use matrix-vector operation method between layer, projection layer and output layer, the leaf node in output layer can all be produced due to branch
A raw probability, these probability multiplications can be obtained by the probability of related phrase, wherein by the general of one phrase of context-prediction
The calculation method of rate is as follows:
P(wi| Context)=P (wi|wi-k, wi-k+1..., wi-1, wi+1..., wi+k)
In above formula, P is phrase probability, wiIndicate some word in text.
Step 3: semantic mood analysis.The calculating of mood value is carried out to the mood keyword that step 2c is extracted, and then is obtained
The mood value of entire text.Specific step is as follows:
Step 3a: mood keyword mood value calculates.Mood value, feelings are assigned to the mood keyword extracted in step 2c
Thread value includes emotional valence and emotional intensity, and emotional valence point is positive and negative, and positive positive expectations and neutral emotional valence are "+", is disappeared
Extremely panic emotional valence is "-", and emotional intensity represents corresponding numerical value.
Emotional intensity is by specific mood keyword according to the assignment rule value of table 1:
Table 1: the assignment rule of mood keyword
The mood value of each mood keyword in text is obtained by the assignment rule of table 1.
Step 3b: text mood value calculates.The mood value of keyword be in a bad mood in text is subjected to aggregation summation and is calculated
Mean value out, to the mood value of text.Its specific calculating is as follows:
Assuming that entire text includes n mood keyword, S1、S2......Sn, then
Wherein E (T) indicates the mood value of a text chapter, SiIndicate i-th of mood keyword in text, E (Si) indicate
The mood value of i-th of mood keyword.
Step 4: text mood tendency determines.
It is compared, is determined whole by the obtained text mood value of step 3b and the text mood tendency decision rule of table 2
Whether a text is passive panic mood.
Table 2: text mood is inclined to decision rule
Text mood value | Text mood tendency |
Just | It is actively optimistic or neutral |
It is negative | It is passive panic |
Case study on implementation
The present embodiment selects the article " Lao Wang viewpoint 05/12 (increasing is held plus storehouse) " in Homeway.com's blog as case study on implementation
Investor's fear mood is carried out to estimate.
The blog text is following (mood keyword therein is the word being underlined):
Trading for stock markets of Shanghai and Shenzhen is slightly seen in the past few daysIt gets warm again after a cold spell.Main force's contract forward price of this week stock index IH and IF are distinguishedGo up
3.24% and 3.00%.All amounts of increase of this week Dalian PP, Zhengzhou PTA will be obviousLagIn SY, PVC substantially just not how
Rise.The more cotton corns configured for a long time as in sell the performance of big macroscopical arbitrage portfolio this week of base metalSteadily, equity change
Change little.The emphasis of the more empty concerns in market has returned in the basic side contradiction of industry itself, the downstream cycle culture efficiency of pig price
ContinueIt is sluggishIt is quick with dregs of beansExpansion.In addition, Zhengzhou apple striking a bargain and always holding position in exchange market supervision department
Serious expression in the eyes stares at lower rapidAmplification, Yi Ran becomes the dazzling hot kind of current year agricultural commodity futures kind.As for
The positive set of the tradition between the screw-thread steel RB contract of Shanghai repeatedly referred to, the performance of this weekIt is straight and narrow。
By 8 mood keywords of above-mentioned Text Feature Extraction, be respectively as follows: improvements, rise, lag, steady, sluggish, expansion,
Amplify, is straight and narrow.
Above-mentioned 8 mood keywords are classified according to mood dictionary, obtain mood value as shown in table 3.
Table 3: the mood value that each mood keyword root of this article is obtained according to mood keyword assignment rule
The mood value of this text calculates as follows according to formula (1):
It is 0.325 according to the text mood value that the present embodiment is calculated in formula (1), according to text mood tendency judgement rule
Then, the emotional valence of the present embodiment is "+", is not belonging to panic mood.
Present embodiments provide a kind of large agricultural product investor fear mood Measurement Method based on semantic text.It should
Method is acquired and is divided by some text informations relevant to large market for farm products issued to investor on internet
Analysis, the text information of acquisition is pre-processed, extract mood keyword go forward side by side market thread value calculating, to obtain entire text
Mood value, and then judgment rule is inclined to by text mood, text mood tendency is determined, to realize investor's fear
The purpose that mood is estimated.
The embodiments of the present invention described above are not intended to limit the scope of the present invention.It is any in the present invention
Spirit and principle within made modifications, equivalent substitutions and improvements etc., should be included in claim protection model of the invention
Within enclosing.
Claims (3)
1. large agricultural product investor fear mood Measurement Method based on semantic text, comprising the following steps:
(1) investment text information acquisition: investing text information library by establishing, major from internet using the form of web crawlers
Related text content is acquired in ancestor's agricultural product investment forum, blog;
(2) text information pre-processes: after investment text information library is established, needing to pre-process text information, pretreatment packet
Punctuate, participle are included, mood keyword extraction is prepared for subsequent semantic mood analysis;
(3) semantic mood analysis: the calculating of mood value is carried out to the mood keyword that step (2) is extracted, and then obtains entire text
This mood value;
The mood value is made of emotional valence and emotional intensity:
The emotional valence be divided by mood tendency it is positive and negative, wherein positive positive expectations and the emotional valence of neutral mood are "+",
The emotional valence of passive fear mood is "-";
The emotional intensity is by specific mood keyword according to the assignment rule value of table 1:
Table 1: the assignment rule of mood keyword
Obtain the mood value of each mood keyword in text by the assignment rule of table 1, and then by pass of being in a bad mood in text
The mood value of keyword carries out aggregation summation and calculates mean value, obtains the mood value of entire text, calculation formula is as follows:
Wherein E (T) indicates the mood value of a text chapter, SiIndicate i-th of mood keyword in text, E (Si) indicate i-th
The mood value of a mood keyword, n indicate the quantity of mood keyword in text;
(4) text mood tendency determines:
Table 2: text mood is inclined to decision rule
It is compared by the obtained text mood value of step (3) and the text mood tendency decision rule of table 2, determines entire text
Whether this is passive panic mood.
2. large agricultural product investor fear mood Measurement Method according to claim 1, which is characterized in that in step
(1) in, investment text information acquisition specifically includes the following steps:
(5) target topic and target acquisition web page interlinkage are determined;Comment is invested according to large agricultural product investor of user's input
Keyword obtains relevant forum, blog network address as target and acquires website as target topic, automatic search;It is acquired in target
In multiple web page interlinkages that website includes, the corresponding target webpage link of the target topic is determined;
(6) processing is filtered to determining target acquisition web page interlinkage;It, may packet in determining target acquisition web page interlinkage
Containing repetition, invalid web page interlinkage, need to be filtered processing;
(7) web page contents downloading is carried out to the target webpage link after filtration treatment, it is fixed according to the html tag in web page contents
Position needs the corresponding URL of article, the corresponding URL of the article acquired as needed that acquire, to the corresponding text information of article into
Row is downloaded and is saved in investment text information library.
3. large agricultural product investor fear mood Measurement Method according to claim 1, which is characterized in that in step
(2) in, text information pretreatment specifically includes the following steps:
(8) punctuate is handled: using Chinese fullstop or other punctuation marks as punctuate node, being needed before carrying out semantic mood and analyzing pair
Collected text information carries out punctuate processing;
(9) participle and part-of-speech tagging: the text after punctuate is segmented, is segmented during participle using Python3.0
The foundation of software, participle is Chinese corpus dictionary, using the HanLP natural language processing packet of open source, in the process of participle
In be automatically performed part-of-speech tagging, the result of participle is each phrase;
(10) it extracts mood keyword and constructs mood dictionary: for convenience of mood keyword extraction, needing to construct mood dictionary, pass through
To all document mood phrases that participle in step (9) and part-of-speech tagging obtain, these phrases are counted, are obtained in network data
There is the highest phrase of word frequency as mood keyword;Mood dictionary is added in the mood keyword of extraction, and subsequent continuous
Expand;Phrase after segmenting in text is compared with the phrase in mood dictionary, comparing successfully is mood keyword,
Mood keyword is extracted;
The mood dictionary creation is using CBOW in word2vector model, including input layer, projection layer, output layer;It is wherein defeated
Entering layer is 2c term vector in context (w), V (Context (w)1)、V(Context(w)2)…V(Context(w)2c);It throws
Shadow layer is then the sum of adding up for 2c term vector of input layer, and the corresponding binary tree of output layer, it is to occur in text
Word as leaf node, the number occurred in the text with each word works as the Huffman tree that constructs of weight;Input layer, throwing
Matrix-vector operation method is used between shadow layer and output layer, the leaf node in output layer can all generate one generally due to branch
Rate, these probability multiplications can be obtained by the probability of related phrase, wherein the calculating of the probability by one phrase of context-prediction
Method is as follows:
P(wi| Context)=P (wi|wi-k, wi-k+1..., wi-1, wi+1..., wi+k)
In above formula, P is phrase probability, wiIndicate some word in text.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811009600.9A CN109145302A (en) | 2018-08-30 | 2018-08-30 | Large agricultural product investor fear mood Measurement Method based on semantic text |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811009600.9A CN109145302A (en) | 2018-08-30 | 2018-08-30 | Large agricultural product investor fear mood Measurement Method based on semantic text |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109145302A true CN109145302A (en) | 2019-01-04 |
Family
ID=64825843
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811009600.9A Withdrawn CN109145302A (en) | 2018-08-30 | 2018-08-30 | Large agricultural product investor fear mood Measurement Method based on semantic text |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109145302A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110400173A (en) * | 2019-07-23 | 2019-11-01 | 中译语通科技股份有限公司 | Market sentiment monitoring system method for building up and system |
CN113450793A (en) * | 2021-06-25 | 2021-09-28 | 平安科技(深圳)有限公司 | User emotion analysis method and device, computer readable storage medium and server |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012134180A2 (en) * | 2011-03-28 | 2012-10-04 | 가톨릭대학교 산학협력단 | Emotion classification method for analyzing inherent emotions in a sentence, and emotion classification method for multiple sentences using context information |
CN103299330A (en) * | 2010-10-21 | 2013-09-11 | 圣脑私营有限责任公司 | Method and apparatus for neuropsychological modeling of human experience and purchasing behavior |
CN106095777A (en) * | 2016-05-26 | 2016-11-09 | 优品财富管理有限公司 | The many empty sentiment indicator methods of prediction securities markets based on big data |
CN108108433A (en) * | 2017-12-19 | 2018-06-01 | 杭州电子科技大学 | A kind of rule-based and the data network integration sentiment analysis method |
-
2018
- 2018-08-30 CN CN201811009600.9A patent/CN109145302A/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103299330A (en) * | 2010-10-21 | 2013-09-11 | 圣脑私营有限责任公司 | Method and apparatus for neuropsychological modeling of human experience and purchasing behavior |
WO2012134180A2 (en) * | 2011-03-28 | 2012-10-04 | 가톨릭대학교 산학협력단 | Emotion classification method for analyzing inherent emotions in a sentence, and emotion classification method for multiple sentences using context information |
CN106095777A (en) * | 2016-05-26 | 2016-11-09 | 优品财富管理有限公司 | The many empty sentiment indicator methods of prediction securities markets based on big data |
CN108108433A (en) * | 2017-12-19 | 2018-06-01 | 杭州电子科技大学 | A kind of rule-based and the data network integration sentiment analysis method |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110400173A (en) * | 2019-07-23 | 2019-11-01 | 中译语通科技股份有限公司 | Market sentiment monitoring system method for building up and system |
CN113450793A (en) * | 2021-06-25 | 2021-09-28 | 平安科技(深圳)有限公司 | User emotion analysis method and device, computer readable storage medium and server |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105808526B (en) | Commodity short text core word extracting method and device | |
CN108733816B (en) | Microblog emergency detection method | |
Batool et al. | Precise tweet classification and sentiment analysis | |
CN105117428B (en) | A kind of web comment sentiment analysis method based on word alignment model | |
Chowdhury et al. | FBK-irst: A multi-phase kernel based approach for drug-drug interaction detection and classification that exploits linguistic information | |
CN106294425B (en) | The automatic image-text method of abstracting and system of commodity network of relation article | |
Amancio et al. | Comparing intermittency and network measurements of words and their dependence on authorship | |
US8713028B2 (en) | Related news articles | |
CN105975453A (en) | Method and device for comment label extraction | |
CN104268197A (en) | Industry comment data fine grain sentiment analysis method | |
CN103207855A (en) | Fine-grained sentiment analysis system and method specific to product comment information | |
CN110931128B (en) | Method, system and device for automatically identifying unsupervised symptoms of unstructured medical texts | |
CN102682120A (en) | Method,device and system for acquiring essential article commented on network | |
CN104391969B (en) | Determine the method and device of user's query statement syntactic structure | |
CN108363699A (en) | A kind of netizen's school work mood analysis method based on Baidu's mhkc | |
CN111026965A (en) | Hot topic tracing method and device based on knowledge graph | |
CN109145302A (en) | Large agricultural product investor fear mood Measurement Method based on semantic text | |
CN106202299A (en) | A kind of people with disability authority user based on people with disability's feature recommends method | |
Pochampally et al. | Semi-supervised automatic generation of wikipedia articles for named entities | |
CN108021657A (en) | A kind of similar author's searching method based on document title semantic information | |
CN104063422B (en) | The feature dictionary iteration update method and device in field in social networks | |
CN104679836B (en) | A kind of automatic extending method of Agricultural ontology | |
KR20140044162A (en) | Duplication news detection system and method for detecting duplication news | |
CN110110079A (en) | A kind of social networks junk user detection method | |
Rouhizadeh et al. | Collecting semantic data from mechanical Turk for a lexical knowledge resource in a text to picture generating system |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190104 |
|
WW01 | Invention patent application withdrawn after publication |