CN111460158A - Microblog topic public emotion prediction method based on emotion analysis - Google Patents
Microblog topic public emotion prediction method based on emotion analysis Download PDFInfo
- Publication number
- CN111460158A CN111460158A CN202010250842.8A CN202010250842A CN111460158A CN 111460158 A CN111460158 A CN 111460158A CN 202010250842 A CN202010250842 A CN 202010250842A CN 111460158 A CN111460158 A CN 111460158A
- Authority
- CN
- China
- Prior art keywords
- emotion
- words
- word
- text
- sentence
- 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
- 230000008451 emotion Effects 0.000 title claims abstract description 147
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000004458 analytical method Methods 0.000 title claims abstract description 27
- 230000003068 static effect Effects 0.000 claims abstract description 13
- 238000013145 classification model Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000002996 emotional effect Effects 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 9
- 230000011218 segmentation Effects 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 150000001875 compounds Chemical class 0.000 claims 1
- 238000007619 statistical method Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 6
- 239000003607 modifier Substances 0.000 description 6
- 230000006399 behavior Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000007935 neutral effect Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
Images
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
- G06F16/35—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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Machine Translation (AREA)
Abstract
The invention provides a microblog topic public sentiment prediction method based on sentiment analysis, which comprises the following steps of: step 1: preprocessing the text; step 2: expanding the basic emotion dictionary by adopting an SO-PMI algorithm; and step 3: constructing a static emotion classification model; and 4, step 4: and constructing an emotion prediction model based on the conditional random field. The method adopts the method of expanding the emotion dictionary to carry out static emotion analysis, improves the accuracy of emotion analysis, introduces the conditional random field to construct the emotion prediction model, carries out public emotion prediction, and is convenient for the government to supervise public opinion and for enterprises to evaluate products in the market.
Description
Technical Field
The invention relates to the field of natural language processing, in particular to a microblog topic public sentiment prediction method based on sentiment analysis.
Background
In the internet era, various social communication software is abundant, explosive information is enriched on the network, people release their own opinions through the network, and the information has a certain social benefit and economic benefit. For governments, understanding the opinions of the masses and paying attention to public opinion guidance in time have great significance for social management; for enterprises, strategic adjustment can be made in time by paying attention to potential consumer feedback and market tendency when products are put on the market, so that economic benefits are improved.
Currently, the research of emotion analysis is mainly inclined to static emotion analysis, and the emotion inclination of a single text, a space or a group is researched. The adopted research methods are mainly divided into two methods, namely an emotion analysis method based on an emotion dictionary and a method based on machine learning. For the emotion analysis method based on the emotion dictionary, the key of the emotion analysis accuracy rate is the quality of the emotion dictionary, and for the emotion analysis method based on machine learning, feature selection is mainly performed. And extracting the features of the analyzed text, enabling the model to learn the text features on the training set in a machine learning mode, and classifying the target text. The judgment of the static emotion analysis is to describe the basis of dynamic emotion propagation, expand a basic emotion dictionary, establish an effective static emotion classification model, extract the characteristics of influence on emotion conversion and use a conditional random field to establish an emotion propagation prediction model.
Disclosure of Invention
The invention aims to provide a topic public emotion prediction method, which improves the accuracy of static emotion analysis by expanding a basic emotion dictionary, and on the basis, uses a conditional random field to construct an emotion prediction propagation model for further predicting the emotion propagation of a topic.
The invention adopts the following technical scheme for realizing the purpose:
a microblog topic public sentiment prediction method based on sentiment analysis comprises the following steps:
(1) text preprocessing
And filtering the sentences to be subjected to emotion analysis, breaking the sentences, and performing word segmentation and other operations on each sentence.
(2) Extended basic emotion dictionary
On the basis of the existing Chinese basic emotion dictionary, words in the microblog topic field are used as expansion key points to expand the Chinese emotion dictionary.
(3) Constructing a static emotion classification model
And carrying out sentiment analysis on the sentence space. Traversing the preprocessed text by adopting a method based on an emotion dictionary, if a word belongs to an emotion word, acquiring an emotion value of the emotion word in the emotion dictionary for recording, searching whether the word has a corresponding degree adverb, a negative word and other modifiers, and if the modifier exists, acquiring a corresponding weight. And finally, weighting and summing the scores to obtain a final score. If the score is positive, judging the emotional state to be positive; if the score is negative, judging the emotional state as negative; if the value is equal to zero, the emotional state has no obvious tendency, and the result is judged to be neutral.
(4) Emotion dynamic propagation prediction based on conditional random field
Obtaining a path of information in a propagation process for modeling, carrying out emotion judgment on information represented by a previous-level node of a node to be predicted for emotion by using the emotion classification model constructed in the step 3, judging the emotion state of an information body, defining the personal information characteristics of a user, the historical behavior characteristics of the user and the interaction characteristics among users, carrying out model training, wherein the result generated by the trained model is the probability value of the emotion polarity of the next-level propagation of the event.
(5) And obtaining a final emotion prediction result.
The text preprocessing in the step (1) comprises the following specific operations:
(1.1) removing stop words and special characters in the microblog text to be processed, wherein the stop words and the special characters comprise user names mentioned in the microblog text when a user issues a microblog, and @ is used as an initial mark.
And (1.2) carrying out sentence breaking on the plain text according to punctuation marks, and carrying out word segmentation by using a jieba word segmentation tool to obtain the part of speech of the word.
The method for expanding the basic emotion dictionary comprises the following specific steps:
and (2.1) respectively selecting 20 positive words and negative words with obvious emotion tendency from the microblog corpus according to the word frequency statistical result, and carrying out synonym expansion to obtain another 20 words serving as reference words.
And 2.2, acquiring words which do not exist in the basic emotion dictionary from the microblog corpus to serve as candidate emotion words.
And (2.3) if the candidate emotion words exist, performing step1.4, otherwise, finishing updating the emotion dictionary.
(2.4) carrying out point mutual information calculation on the candidate emotion words and the commendative word groups and the derogative word groups which are taken as the reference words in sequence, wherein the calculation formula is as follows:
wherein, N represents the total number of documents in the corpus, Pword represents positive words, Nword represents negative words, df (word) represents the document frequency of the words word in the corpus, and df (word & Pword) represents the document frequency of the common occurrence of the two words in the corpus.
And (2.5) judging the reference word as the corresponding word type (commendability or derogation) according to the calculation result, and updating the reference word into the basic dictionary.
When the static emotion classification model is constructed, the specific steps are as follows:
and (3.1) adopting a method based on extending emotion dictionary in emotion analysis of the text, and taking the weight of the emotion words contained in each sentence in the emotion dictionary as the initial emotion value of the sentence.
And (3.2) setting different weights for different modifier types according to different emotion word modifiers in the sentences, and multiplying the initial emotion value of each sentence by the corresponding weight.
(3.2.1) degree adverbs can influence the expression degree of the sentence emotion, in the patent, common degree adverbs are summarized and divided into four levels, and the weight of each level is defined as a degree adverb table. And multiplying the emotion value by the corresponding weight according to the corresponding degree adverb of the emotion word in the sentence.
(3.2.2) negating adverbs changes the polarity of the sentence emotion. In the patent, common negative adverbs are summarized and defined as-1 in a unified way, and the corresponding weight is multiplied according to whether negative adverb modification exists in the sentence emotional words.
And (3.3) the exclamation sentence can enhance the expression degree of the sentence emotion, and the question sentence can change the polarity of the emotion expression. And dividing sentence types according to punctuation marks. According to different sentence types, different weight values are set, and then the currently obtained emotion value is multiplied by the corresponding weight value.
And (3.4) accumulating the emotion values of each sentence in the text to obtain the emotion state of the whole text.
The emotion dynamic propagation prediction method based on the conditional random field comprises the following specific steps:
and (4.1) modeling a microblog propagation path, and storing node information of each stage of forwarding process. The probability formula of the emotional state of the node to be predicted is written as follows:
wherein f ism(d,si-1,siU) is a characteristic function, λmThe parameter values are corresponding to the characteristic functions, Z (u) is a normalization factor, s represents the emotional state, d represents the emotional state of the original microblog text, and u represents the user sequence participating in forwarding.
(4.2) defining characteristics capable of influencing the propagation of emotional states.
And (4.2.1) acquiring original microblog emotion polarities through a static emotion classification model, wherein the original microblog emotion polarities comprise positive, negative or neutral.
And (4.2.2) acquiring personal information characteristics of the user involved in the forwarding path, including the number of fans, the number of concerns and whether the user is an authenticated user.
And (4.2.3) acquiring historical behavior characteristics of the user, and using the forwarding and original microblog quantity as a standard for judging whether the user is active or not.
And (4.2.4) acquiring the interactive characteristics among the users, the characteristics of the users at the upper level and the emotional state of the microblog text.
(4.3) predicting probability value formula based on conditional random field emotional state to rewrite
Wherein f isDm(d,si) As a function of the original microblog emotional characteristics, fUm(d,si,ui) As a function of characteristics of the subscriber information, fBm(d,si,ui) As a function of the user's historical behavior characteristics, fIm(d,si-1,si,ui-1,ui) As a function of the characteristics of the interaction between the users.
And (4.4) training a prediction model by using the conversion data under the microblog propagation path to realize parameter estimation.
Has the advantages that:
compared with the prior art, the invention has the beneficial effects that:
(1) and the SO-PMI algorithm is used for expanding the emotion dictionary, SO that the number of emotion words of the emotion dictionary is increased, the accuracy of static emotion analysis is improved, and the foundation for dynamic emotion analysis prediction is laid.
(2) And extracting the characteristics influencing the emotion propagation, constructing an emotion propagation prediction model by adopting a conditional random field, analyzing the dynamic process of the emotion propagation, and effectively predicting the emotion tendency of the next propagation node.
Drawings
FIG. 1 is a flowchart of a microblog topic public sentiment prediction method based on sentiment analysis;
FIG. 2 is a flow diagram of a basic emotion dictionary expansion process;
FIG. 3 is a schematic view of a static sentiment analysis flow;
FIG. 4 is a schematic diagram of a dynamic emotion prediction process.
Detailed Description
The invention is further illustrated by the following specific examples.
The first embodiment is as follows: the invention provides a microblog topic public sentiment prediction method based on sentiment analysis, which is shown in figure 1. The method comprises the following specific steps.
S1 extended basic emotion dictionary
And (3) preprocessing the microblog corpus by taking the Dalian Chinese emotion dictionary as a basic emotion dictionary, selecting a reference word, identifying a new word, and calculating point mutual information of the reference word and the new word. And dividing the commendably and derogatory words according to the calculated result and adding the commendably and derogatory words into the emotion dictionary.
The following describes in detail the extension method of the basic emotion dictionary with reference to fig. 2, specifically as follows:
s1.1, performing word segmentation on the text in the corpus by using a jieba word segmentation tool, and performing part-of-speech tagging. And dividing words with parts of speech being verbs, adjectives and adverbs into emotional words, filtering the words existing in the basic emotional dictionary, and taking the screened words as candidate emotional words.
S1.2, counting the frequency of the words in the corpus, arranging the words in a sequence from high to low, screening out the recognition words and the derogation words with higher frequency, and selecting the corresponding synonyms as the reference words.
S1.2.1, in this embodiment, 10 respective positive and negative words with the highest frequency are selected, and the synonyms corresponding to each are obtained through the synonym dictionary, so that a total of 40 reference words are obtained.
S1.2.2, in this example, 10 thousands of microblogs under the social topic theme are crawled as a corpus by crawlers to screen words.
S1.3, according to the reference words screened out by the method, point mutual information values between the candidate emotion words and the reference words are sequentially calculated, and the words are added into an emotion dictionary through the value range.
S1.3.1, in this embodiment, the point mutual information value of each candidate word and the reference word is calculated by the following formula:
if the S-P value is larger than zero, the candidate word is updated to the commendation word dictionary, and if the S-P value is smaller than zero, the candidate word is updated to the derogation word dictionary. And taking the expanded emotion dictionary as an emotion analysis tool.
The emotion analysis process in constructing the emotion analysis model will be described in detail below with reference to fig. 3.
S2.1, according to the emotion words existing in the text to be analyzed, the emotion value of the emotion words in an emotion dictionary is searched and used as the initial emotion value of the sentence.
And S2.2, judging whether the emotion words have modifier modification to form word pairs, and performing corresponding rule calculation according to corresponding modifiers.
S2.2.1, if the combination of the words of the emotion word modified by the negative word is combined, the initial emotion value is multiplied by-1, and the emotion polarity is changed. And if the combination of the words of the emotional words modified by the degree adverb is combined, multiplying the initial emotion value by the weight value under the corresponding level of the degree adverb.
In this patent, the degree adverbs are divided into four levels, low, medium, high and high, and the weights at the corresponding levels are 0.75, 1, 1.25 and 1.5, respectively.
And S2.3, judging the sentence pattern of the sentence according to the punctuation marks. If it is "! ", then multiply the current sentiment value by 2. If so. "end", no processing is done. If with "? If the sentence is a question-back sentence, the user needs to judge whether the sentence is a question-back sentence.
This patent designs a word list of questions, in which are "? If the word in the table exists, the sentence is classified as a question-reversing sentence.
And S2.4, accumulating and summing the sentences endowed with the emotion values to obtain final emotion values.
And summing the emotion values obtained from each sentence of each text to be analyzed, if the value is greater than zero, judging that the text emotion polarity is positive, if the value is less than zero, judging that the text emotion polarity is negative, and if the value is equal to zero, judging that the text emotion polarity is neutral, wherein no obvious emotion tendency exists.
The emotion prediction process is described in detail below with reference to fig. 4, specifically as follows:
and S3.1, modeling a microblog propagation path, and conforming to the conditional random field model. Inputting a microblog text d and an attribute sequence u of a user participating in forwarding, and obtaining a conditional probability formula of emotional state output, wherein the conditional probability formula is written as follows:
s3.2, defining the characteristics influencing the emotion of the next node, including objective emotional tendency of the microblog body, personal information of the user, historical behavior information of the user and interaction information among the users, and rewriting a formula as follows:
and S3.3, training a model and estimating parameters.
S3.3.1, in this embodiment, 10 ten thousand corpora indicating emotional tendency are obtained for model training and parameter estimation.
S3.3.2, wherein the designation of emotional tendencies uses the process and method of the previous step.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (5)
1. A method for predicting public feelings of Chinese microblog topics is characterized by comprising the following steps:
step 1: text preprocessing: filtering special symbols of a text to be analyzed, breaking sentences according to punctuation marks, dividing the sentences into complex sentences and clauses, and performing word segmentation;
step 2: expanding a basic emotion dictionary: adding new emotion words which do not exist in the basic emotion dictionary in the corpus into the basic emotion dictionary by using point mutual information, and widening the use range;
and step 3: constructing a static emotion classification model: setting rules by taking the expanded basic emotion dictionary as a tool, calculating the emotion value of each text segment from the sequence of word pairs to sentences to paragraphs, and constructing a static emotion classification model;
and 4, step 4: carrying out emotion dynamic propagation prediction based on the conditional random field: and modeling by using a conditional random field, constructing a microblog propagation path, and predicting the probability of the emotional tendency of the next propagation node.
2. The method for predicting the public sentiment of the Chinese microblog topics according to claim 1, wherein the method comprises the following steps:
step 1.1: processing the input unstructured text according to a fixed format: removing a non-text part of the acquired unstructured text, and simultaneously removing stop words and a user name (taking an @ symbol as an initial mark) mentioned in the text, thereby acquiring a plain text for the next processing;
step 1.2: according to ". ","? ","! Punctuation marks of "and" … … "are used to make punctuation of text into a sentence, which is a compound sentence, according to", "; "punctuate punctuation with equal punctuation as clauses;
step 1.3: and performing word segmentation on the processed plain text, and performing part-of-speech tagging. And filtering the auxiliary words, prepositions, sound-making words, quantifiers and digital words obtained by word segmentation according to the part of speech.
3. The method for predicting the public sentiment of the Chinese microblog topics according to claim 1, wherein the method comprises the following steps:
step 2.1: marking the part of speech as adjectives, adverbs, nouns and verbs, and taking words which do not exist in the basic emotion dictionary as candidate emotion words;
step 2.2: screening recognition words and depreciation words with the frequency of occurrence of the first ten in a corpus respectively by using a statistical method, and acquiring synonyms of the recognition words and the depreciation words as reference words;
step 2.3: calculating point mutual information values of the candidate emotion words and the reference words in sequence, and adding the point mutual information values to corresponding positions in the basic emotion dictionary according to results;
step 2.4: the mutual information calculation formula of each candidate emotion word and the reference word is as follows:
wherein, N represents the total number of documents in the corpus, Pword represents positive words, Nword represents negative words, df (word) represents the document frequency of the words word in the corpus, and df (word & Pword) represents the document frequency of the common occurrence of the two words in the corpus.
4. The method for predicting the public sentiment of the Chinese microblog topics as claimed in claim 1, wherein the method comprises the following steps:
step 3.1: taking the emotion value of the emotion word in each sentence as the initial emotion value of the sentence;
step 3.2: the emotion values of the word pairs are calculated. If the emotion word is modified by the degree adverb, multiplying the initial emotion value by the corresponding weight, and if the emotion word is modified by the negative adverb, multiplying the initial emotion value by negative 1 to change the emotion polarity;
step 3.3: and calculating the emotion value of the sentence. Multiplying the emotion value of the obtained word pair by the corresponding weight according to different sentence patterns;
step 3.4: and accumulating the emotion values of each sentence in the whole text to obtain the emotion tendency of the whole text.
5. The method for predicting the public sentiment of the Chinese microblog topics according to claim 1, wherein the method comprises the following steps: modeling of a microblog propagation path is carried out based on a conditional random field, characteristics capable of influencing emotional tendency in the propagation process are defined, model training and parameter estimation are carried out, and the method in the step 3 is used in the part related to text emotion analysis.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010250842.8A CN111460158B (en) | 2020-04-01 | 2020-04-01 | Microblog topic public emotion prediction method based on emotion analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010250842.8A CN111460158B (en) | 2020-04-01 | 2020-04-01 | Microblog topic public emotion prediction method based on emotion analysis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111460158A true CN111460158A (en) | 2020-07-28 |
CN111460158B CN111460158B (en) | 2022-09-23 |
Family
ID=71684349
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010250842.8A Active CN111460158B (en) | 2020-04-01 | 2020-04-01 | Microblog topic public emotion prediction method based on emotion analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111460158B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113064989A (en) * | 2020-11-19 | 2021-07-02 | 龚越 | Public transport opinion perception feature module and analysis and treatment application |
CN114118069A (en) * | 2021-09-02 | 2022-03-01 | 四川启睿克科技有限公司 | Emotion dictionary expansion method and emotion polarity analysis method based on SOPMI algorithm |
CN114742281A (en) * | 2022-03-21 | 2022-07-12 | 西北工业大学 | Public event network public opinion popularity prediction method based on grey model |
CN116828246A (en) * | 2023-06-29 | 2023-09-29 | 中科智宏(北京)科技有限公司 | Digital live broadcast interaction method, system, equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080249764A1 (en) * | 2007-03-01 | 2008-10-09 | Microsoft Corporation | Smart Sentiment Classifier for Product Reviews |
CN104731770A (en) * | 2015-03-23 | 2015-06-24 | 中国科学技术大学苏州研究院 | Chinese microblog emotion analysis method based on rules and statistical model |
CN106569996A (en) * | 2016-03-30 | 2017-04-19 | 广东工业大学 | Chinese-microblog-oriented emotional tendency analysis method |
CN110096696A (en) * | 2018-06-11 | 2019-08-06 | 电子科技大学 | A kind of Chinese long text sentiment analysis method |
CN110489523A (en) * | 2019-07-31 | 2019-11-22 | 西安理工大学 | A kind of fine granularity sentiment analysis method based on online shopping evaluation |
-
2020
- 2020-04-01 CN CN202010250842.8A patent/CN111460158B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080249764A1 (en) * | 2007-03-01 | 2008-10-09 | Microsoft Corporation | Smart Sentiment Classifier for Product Reviews |
CN104731770A (en) * | 2015-03-23 | 2015-06-24 | 中国科学技术大学苏州研究院 | Chinese microblog emotion analysis method based on rules and statistical model |
CN106569996A (en) * | 2016-03-30 | 2017-04-19 | 广东工业大学 | Chinese-microblog-oriented emotional tendency analysis method |
CN110096696A (en) * | 2018-06-11 | 2019-08-06 | 电子科技大学 | A kind of Chinese long text sentiment analysis method |
CN110489523A (en) * | 2019-07-31 | 2019-11-22 | 西安理工大学 | A kind of fine granularity sentiment analysis method based on online shopping evaluation |
Non-Patent Citations (2)
Title |
---|
JIAN ZHU: ""CRF Based on LHFS Applied on Sentiment Classification"", 《INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE AND ELECTRONICS ENGINEERING》, 25 March 2012 (2012-03-25) * |
胡召亚 等: ""基于关键句提取的中文微博情感计算"", 《阜阳师范学院学报(自然科学版)》, vol. 36, no. 3, 30 September 2019 (2019-09-30) * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113064989A (en) * | 2020-11-19 | 2021-07-02 | 龚越 | Public transport opinion perception feature module and analysis and treatment application |
CN113064989B (en) * | 2020-11-19 | 2022-12-02 | 龚越 | Method for extracting perception features of public transport sentiments |
CN114118069A (en) * | 2021-09-02 | 2022-03-01 | 四川启睿克科技有限公司 | Emotion dictionary expansion method and emotion polarity analysis method based on SOPMI algorithm |
CN114742281A (en) * | 2022-03-21 | 2022-07-12 | 西北工业大学 | Public event network public opinion popularity prediction method based on grey model |
CN116828246A (en) * | 2023-06-29 | 2023-09-29 | 中科智宏(北京)科技有限公司 | Digital live broadcast interaction method, system, equipment and storage medium |
CN116828246B (en) * | 2023-06-29 | 2024-03-19 | 中科智宏(北京)科技有限公司 | Digital live broadcast interaction method, system, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN111460158B (en) | 2022-09-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111460158B (en) | Microblog topic public emotion prediction method based on emotion analysis | |
CN108052593B (en) | Topic keyword extraction method based on topic word vector and network structure | |
CN108287922B (en) | Text data viewpoint abstract mining method fusing topic attributes and emotional information | |
RU2662688C1 (en) | Extraction of information from sanitary blocks of documents using micromodels on basis of ontology | |
US7873584B2 (en) | Method and system for classifying users of a computer network | |
CN112749341B (en) | Important public opinion recommendation method, readable storage medium and data processing device | |
Ginting et al. | Hate speech detection on twitter using multinomial logistic regression classification method | |
Kaushik et al. | A study on sentiment analysis: methods and tools | |
WO2024131111A1 (en) | Intelligent writing method and apparatus, device, and nonvolatile readable storage medium | |
JP2005122533A (en) | Question-answering system and question-answering processing method | |
Lu et al. | Sentiment analysis of film review texts based on sentiment dictionary and SVM | |
CN111626050B (en) | Microblog emotion analysis method based on expression dictionary and emotion general knowledge | |
CN110598219A (en) | Emotion analysis method for broad-bean-net movie comment | |
CN112699240A (en) | Intelligent dynamic mining and classifying method for Chinese emotional characteristic words | |
Amali et al. | Classification of cyberbullying Sinhala language comments on social media | |
Niu et al. | An Improved Method for Web Text Affective Cognition Computing Based on Knowledge Graph. | |
Jaradat et al. | Hybrid-based Arabic single-document text summarization approach using genatic algorithm | |
CN109710762B (en) | Short text clustering method integrating multiple feature weights | |
CN116244446A (en) | Social media cognitive threat detection method and system | |
CN114722176A (en) | Intelligent question answering method, device, medium and electronic equipment | |
Wei et al. | Online education recommendation model based on user behavior data analysis | |
US11605004B2 (en) | Method and system for generating a transitory sentiment community | |
CN104794209A (en) | Chinese microblog sentiment classification method and system based on Markov logic network | |
CN112182159B (en) | Personalized search type dialogue method and system based on semantic representation | |
CN111008285B (en) | Author disambiguation method based on thesis key attribute network |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |