CN110807315A - Topic model-based online comment emotion mining method - Google Patents
Topic model-based online comment emotion mining method Download PDFInfo
- Publication number
- CN110807315A CN110807315A CN201910975438.4A CN201910975438A CN110807315A CN 110807315 A CN110807315 A CN 110807315A CN 201910975438 A CN201910975438 A CN 201910975438A CN 110807315 A CN110807315 A CN 110807315A
- Authority
- CN
- China
- Prior art keywords
- text
- emotion
- distribution
- comment
- pair
- 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
- 230000008451 emotion Effects 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000005065 mining Methods 0.000 title claims abstract description 23
- 239000011159 matrix material Substances 0.000 claims abstract description 22
- 230000002996 emotional effect Effects 0.000 claims abstract description 10
- 238000009826 distribution Methods 0.000 claims description 59
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000012821 model calculation Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 abstract 1
- 230000007547 defect Effects 0.000 description 2
- 238000012856 packing Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an online comment emotion mining method based on a topic model. The method comprises the following specific steps: (1) inputting an online comment text set in any field; (2) extracting text aspect opinion pairs for each text from the comment text set, and constructing a text-aspect opinion pair matrix; (3) constructing an emotional theme model; (4) inputting the text-aspect opinions into the emotion theme model for training; (5) and outputting a comment text emotion mining result. The method adopts the method of extracting the comment text aspect opinion pair to replace the traditional vocabulary as the theme model input, solves the problems of mixed attribute words and viewpoint words and single vocabulary emotion fuzzy, and improves the accuracy and interpretability of comment text emotion analysis; meanwhile, the dimensionality of text representation is reduced, and the model calculation time is reduced; the method is simple and easy to operate and has good effect.
Description
Technical Field
The invention relates to the field of event extraction in information extraction, in particular to an online comment emotion mining method based on a topic model.
Background
Currently, many researchers analyze text emotion by a method of expanding a topic model (LDA) by combining emotion layers, and represent the text as a bag of words, and the bag of words is composed of words in the text. The topic model based on LDA can effectively mine the emotion of the text to a certain extent.
When the traditional topic model extension mode is used for mining the text emotion, the following defects exist: (1) the text is regarded as a bag of words, and the dependency relationship among the words is ignored, so that a great deal of text semantics is lost; (2) the text is represented as a collection of words, making the model input dimension too high and computationally time consuming.
Disclosure of Invention
The invention aims to provide an online comment emotion mining method based on a topic model, aiming at the defect that the traditional topic model expansion method is used for mining text emotion. By adopting the method for extracting the comment text aspect opinion pair to replace the traditional vocabulary as the topic model extension model for input, the problems of mixed attribute words and viewpoint words and single vocabulary emotion fuzzy are solved, and the accuracy and the interpretability of comment text emotion analysis are improved; meanwhile, the dimensionality of text representation is reduced, and the model calculation time is shortened.
In order to achieve the above object, the present invention is conceived as follows: the aspect opinion pair of the comment text is adopted to replace the traditional vocabulary as the input of the topic model, the dimensionality of text representation is reduced, the semantic property of text representation is increased, and more semantic information is provided for the input of the model.
According to the above inventive idea, the invention adopts the following technical scheme:
an online comment emotion mining method based on a topic model comprises the following specific steps:
(1) inputting an online comment text set in any field;
(2) extracting text aspect opinion pairs for each text from the comment text set, and constructing a text-aspect opinion pair matrix;
(3) constructing an emotional theme model;
(4) inputting the text-aspect opinions into the emotion theme model for training;
(5) and outputting a comment text emotion mining result.
The process of extracting the text aspect opinion pair in the step (2) is as follows:
(2-1) dividing each comment text in the comment text set into comment units according to punctuation marks, segmenting each comment unit into words, parts of speech tagging and stop words by adopting a word segmentation tool, wherein adverbs are generally used for modifying adjectives or verbs, so that the adverbs and modified parts of the adverbs are combined into a whole, and then the part of speech of the whole is annotated as the part of speech of the modified word;
(2-2) carrying out mode matching on each comment unit according to the part of speech matching mode, extracting aspect opinion pairs, converting each comment text into a set of aspect opinion pairs, and forming an aspect opinion pair word pair table;
(2-3) calculating the association strength of each item in the aspect opinion paired word pair table by using an association strength calculation formula, arranging the aspect opinion paired items in a descending order according to the association strength, and taking TopN aspect opinion paired items to form a text-aspect opinion paired matrix; the correlation strength calculation formula of the aspect opinion pair is as follows:
wherein, Co (wi, wj) represents the Co-occurrence number of the keywords wi and wj in the comment text extraction unit set, and df (wi) and df (wj) represent the word frequency of the keywords wi and wj in the aspect opinion pair set respectively.
The emotional theme model in the step (3) is constructed as follows:
(3-1) generating a polynomial parameter theta of the topic distribution of the corpus based on the dirichlet hyper-parameter α, wherein α is a parameter of the dirichlet distribution to which theta obeys;
(3-2) generating polynomial parameters of e-th emotion word pair distribution of kth theme of corpus based on Dirichlet hyper-parameter βWherein β is
The obeyed Dirichlet distribution parameters, k and e are positive integers;
(3-3) generating polynomial parameter pi of emotion distribution of kth topic of corpus based on Dirichlet hyper parameter gammak(ii) a Wherein gamma is pikParameters of the obeyed dirichlet distribution;
(3-4) generating topic z of the nth aspect opinion pair of the d document based on the polynomial parameter theta of topic distribution of the corpusdnWhere θ is zdnD and n are positive integers according to the parameters of the polynomial distribution obeyed;
(3-5) polynomial parameter pi based on emotion distribution of kth topic of corpuskAnd subject z of the nth aspect opinion pair of the d documentdnI.e., (pi)k，zdn) Generating sentiment s of nth aspect opinion pair of the d documentdnIn which pizdnIs sdnParameters of the polynomial distribution obeyed;
(3-6) polynomial parameters of e-th emotion lower word pair distribution based on k-th theme of corpus
Topic z of the nth aspect opinion pair of the d documentdnAnd the sentiment s of the nth aspect opinion pair of the d documentdnI.e., (i)zdn，sdn) Generating an nth aspect opinion pair wp of a d documentdnWherein
Is a parameter of the polynomial distribution to which word pairs wp are obeyed;
and (3-7) looping the steps (3-4), (3-5) and (3-6) until all the documents in the corpus are generated.
The step (4) of inputting the text-aspect opinions into the emotion theme model for training comprises the following specific steps:
(4-1) respectively sampling a theme and an emotion for each item in the text-aspect opinion pair matrix, and iterating the process for multiple times to finally form a text-theme matrix and a text-emotion matrix;
(4-2) calculating text-theme probability distribution and theme-emotion probability distribution according to the values of the text-theme matrix and the text-emotion matrix, wherein the specific calculation formula is as follows:
wherein N iskRepresenting the number of word pairs in the text assigned to topic k, N representing the number of word pairs for all of the text set, NkeRepresenting the number of word pairs simultaneously assigned to a subject K and an emotion E, α and gamma representing hyper-parameters of conjugate prior Dirichlet distribution of theta and Π, respectively, E being the number of emotion categories, and K being the number of hidden subjects;
(4-3) further calculating to obtain the overall emotion distribution condition according to the text-theme probability distribution and the theme-emotion probability distribution, wherein the calculation formula is as follows:
R＝θk*Πke
wherein, R represents the sentiment distribution of the online comment collection and is an E-dimensional vector.
Compared with the prior art, the invention has the following outstanding characteristics and advantages:
the method improves the input of the sentiment theme model, and integrates more semantics for the input, thereby improving the accuracy of online comment text sentiment mining results; the text is represented in the form of the face suggestion pair, so that the dimension of model input is reduced, and the model solving time is greatly shortened; the rich semantic information obtained by adding the interword relationship improves the interpretability of the emotion mining result, and is convenient for human to understand the emotion mining result.
Drawings
FIG. 1 is a flowchart of an online comment emotion mining method based on a topic model according to the present invention.
FIG. 2 is an emotional topic model constructed based on the LDA topic model.
Detailed Description
The embodiments of the present invention will be further described with reference to the accompanying drawings.
According to the online comment sentiment mining method based on the topic model, by taking online comments of drinks in the food industry as an example, a certain online comment text set of drinks from 1 month in 2017 to 12 months in 2017 is acquired from a Taobao website for sentiment mining. As shown in fig. 1, the online comment emotion mining method based on the topic model of this embodiment includes the following steps:
s1, inputting a product online comment text set, for example, 10000 comment text sets of a certain drink;
s2, extracting the aspect opinions based on the part of speech matching mode comprises the following steps:
s2.1, extracting text opinion pairs from a comment text set for each text, dividing each comment text in the comment text data set into comment units according to punctuation coincidence, and adopting a word segmentation tool to segment, label and remove stop words for each comment unit, wherein adverbs are generally used for modifying adjectives or verbs, so that the adverbs and modified parts of the adverbs are combined into a whole, and then the part of speech of the whole is annotated as the part of speech of a modified word;
s2.2, performing mode matching on each comment unit according to the part-of-speech matching mode, extracting aspect opinion pairs, converting each comment text into a set of aspect opinion pairs, and forming an aspect opinion pair word pair table;
the part-of-speech matching patterns are as follows:
| matching patterns | Examples of the invention | Extracting the result |
| n.+a. | Taste (non) very good (adv.) good (adj.) | Taste-very good |
| n.+v. | Price (noun) rising (verb) | Price-rise |
| a.+n. | Lovely (adj.) packing box (non) | Packing box-lovely |
| v.+n. | Like (verb) display screen (nun) | Display screen-like |
| n.+a.+v. | Easy (adj.) break (verb) of packaging (noun) | Packaging-easy to break |
| n.+v.+a. | Seller (non) delivery (verb) express (adj.) | Seller-shipping express |
S2.3, calculating the association strength of each item in the aspect opinion pair table by using an association strength calculation formula, arranging the aspect opinion pair items in a descending order according to the association strength, and taking TopN aspect opinion pair items to form a text-aspect opinion pair matrix; the correlation strength calculation formula of the aspect opinion pair is as follows:
wherein, Co (wi, wj) represents the Co-occurrence number of the keywords wi and wj in the comment text extraction unit set, and df (wi) and df (wj) represent the word frequency of the keywords wi and wj in the aspect opinion pair set respectively.
S3, constructing an emotional theme model as shown in FIG. 2, wherein α, β and gamma are Dirichlet super-parameters, theta is corpus theme distribution, pi is emotional theme joint distribution,
the method comprises the following specific steps of distributing word pairs under an emotional theme, wherein K represents the total number of themes, E represents the total number of emotions, N represents the total number of each text word pair, D represents the total number of texts, s represents the hidden variable emotion, z represents the hidden variable theme, and wp represents the observation variable aspect opinion pair, and the specific construction steps are as follows:
s3.1, generating a polynomial parameter theta of the topic distribution of the corpus based on Dirichlet hyper-parameter α, wherein α is the parameter of the Dirichlet distribution to which theta obeys;
s3.2 generating polynomial parameters of e-th emotion lower word pair distribution of kth theme of corpus based on Dirichlet hyper-parameter βWherein β is
The obeyed Dirichlet distribution parameters, k and e are positive integers;
s3.3 generating polynomial parameters pi of emotional distribution of kth topic of corpus based on Dirichlet hyper-parameter gammak(ii) a Wherein gamma is pikParameters of the obeyed dirichlet distribution;
s3.4 generating the d document based on the polynomial parameter theta of the topic distribution of the corpusSubject z of the nth aspect opinion pairdnWhere θ is zdnD and n are positive integers according to the parameters of the polynomial distribution obeyed;
s3.5 polynomial parameter π based on emotional distribution of the kth topic of the corpuskAnd subject z of the nth aspect opinion pair of the d documentdnI.e., (pi)k，zdn) Generating sentiment s of nth aspect opinion pair of the d documentdnIn which pizdnIs sdnParameters of the polynomial distribution obeyed;
s3.6 polynomial parameters of e-th emotion lower word pair distribution based on kth theme of corpus
Topic z of the nth aspect opinion pair of the d documentdnAnd the sentiment s of the nth aspect opinion pair of the d documentdnI.e., (i)
zdn，sdn) Generating an nth aspect opinion pair wp of a d documentdnWherein
Wp is a parameter of the polynomial distribution to which it is subject;
s3.7 loops through steps S3.4, S3.5, S3.6 above until all documents in the corpus are generated.
S4, inputting the text-aspect opinions into the emotion theme model for training, and specifically comprising the following steps:
s4.1, respectively sampling a theme and an emotion for each item in the text-aspect opinion pair matrix, and iterating the process for multiple times to finally form a text-theme matrix and a text-emotion matrix;
s4.2, calculating text-theme probability distribution and theme-emotion probability distribution according to the values of the text-theme matrix and the text-emotion matrix, wherein the specific calculation formulas are as follows:
wherein N iskRepresenting the number of word pairs in the text assigned to topic k, N representing the number of word pairs for all of the text set, NkeRepresenting the number of word pairs simultaneously assigned to a subject K and an emotion E, α and gamma representing hyper-parameters of conjugate prior Dirichlet distribution of theta and Π, respectively, E being the number of emotion categories, and K being the number of hidden subjects;
s4.3, further calculating to obtain the overall emotion distribution condition according to the text-theme probability distribution and the theme-emotion probability distribution, wherein the calculation formula is as follows:
R＝θk*Πke
wherein, R represents the sentiment distribution of the online comment collection and is an E-dimensional vector.
And S5, outputting a comment text emotion mining result.
Claims (4)
1. An online comment emotion mining method based on a topic model is characterized by comprising the following specific steps:
(1) inputting an online comment text set in any field;
(2) extracting text aspect opinion pairs for each text from the comment text set, and constructing a text-aspect opinion pair matrix;
(3) constructing an emotional theme model;
(4) inputting the text-aspect opinions into the emotion theme model for training;
(5) and outputting a comment text emotion mining result.
2. The method for mining online comment emotion based on topic model according to claim 1, wherein the process of extracting text aspect opinion pairs in step (2) is as follows:
(2-1) dividing each comment text in the comment text set into comment units according to punctuation marks, segmenting each comment unit into words, parts of speech tagging and stop words by adopting a word segmentation tool, wherein adverbs are generally used for modifying adjectives or verbs, so that the adverbs and modified parts of the adverbs are combined into a whole, and then the part of speech of the whole is annotated as the part of speech of the modified word;
(2-2) carrying out mode matching on each comment unit according to the part of speech matching mode, extracting aspect opinion pairs, converting each comment text into a set of aspect opinion pairs, and forming an aspect opinion pair word pair table;
(2-3) calculating the association strength of each item in the aspect opinion paired word pair table by using an association strength calculation formula, arranging the aspect opinion paired items in a descending order according to the association strength, and taking TopN aspect opinion paired items to form a text-aspect opinion paired matrix; the correlation strength calculation formula of the aspect opinion pair is as follows:
wherein, Co (wi, wj) represents the Co-occurrence number of the keywords wi and wj in the comment text extraction unit set, and df (wi) and df (wj) represent the word frequency of the keywords wi and wj in the aspect opinion pair set respectively.
3. The topic model-based online comment emotion mining method of claim 1, wherein the emotion topic model of step (3) is constructed by the following process:
(3-1) generating a polynomial parameter theta of the topic distribution of the corpus based on the dirichlet hyper-parameter α, wherein α is a parameter of the dirichlet distribution to which theta obeys;
(3-2) generating polynomial parameters of e-th emotion word pair distribution of kth theme of corpus based on Dirichlet hyper-parameter β
Wherein β is
The obeyed Dirichlet distribution parameters, k and e are positive integers;
(3-3) generating a plurality of emotion distributions of kth topic of corpus based on Dirichlet hyper-parameter gammaParameter of the polynomial pik(ii) a Wherein gamma is pikParameters of the obeyed dirichlet distribution;
(3-4) generating topic z of the nth aspect opinion pair of the d document based on the polynomial parameter theta of topic distribution of the corpusdnWhere θ is zdnD and n are positive integers according to the parameters of the polynomial distribution obeyed;
(3-5) polynomial parameter pi based on emotion distribution of kth topic of corpuskAnd subject z of the nth aspect opinion pair of the d documentdnI.e., (pi)k，zdn) Generating sentiment s of nth aspect opinion pair of the d documentdnIn which pizdnIs sdnParameters of the polynomial distribution obeyed;
(3-6) polynomial parameters of e-th emotion lower word pair distribution based on k-th theme of corpusTopic z of the nth aspect opinion pair of the d documentdnAnd the sentiment s of the nth aspect opinion pair of the d documentdnI.e., (i)
zdn，sdn) Generating an nth aspect opinion pair wp of a d documentdnWherein
Is a parameter of the polynomial distribution to which word pairs wp are obeyed;
and (3-7) looping the steps (3-4), (3-5) and (3-6) until all the documents in the corpus are generated.
4. The topic model-based online comment emotion mining method of claim 1, wherein the text-aspect opinion input matrix in step (4) is trained to the emotion topic model by the specific steps of:
(4-1) respectively sampling a theme and an emotion for each item in the text-aspect opinion pair matrix, and iterating the process for multiple times to finally form a text-theme matrix and a text-emotion matrix;
(4-2) calculating text-theme probability distribution and theme-emotion probability distribution according to the values of the text-theme matrix and the text-emotion matrix, wherein the specific calculation formula is as follows:
wherein N iskRepresenting the number of word pairs in the text assigned to topic k, N representing the number of word pairs for all of the text set, NkeRepresenting the number of word pairs simultaneously assigned to a subject K and an emotion E, α and gamma representing hyper-parameters of conjugate prior Dirichlet distribution of theta and Π, respectively, E being the number of emotion categories, and K being the number of hidden subjects;
(4-3) further calculating to obtain the overall emotion distribution condition according to the text-theme probability distribution and the theme-emotion probability distribution, wherein the calculation formula is as follows:
R＝θk*Πke
wherein, R represents the sentiment distribution of the online comment collection and is an E-dimensional vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910975438.4A CN110807315A (en) | 2019-10-15 | 2019-10-15 | Topic model-based online comment emotion mining method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910975438.4A CN110807315A (en) | 2019-10-15 | 2019-10-15 | Topic model-based online comment emotion mining method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110807315A true CN110807315A (en) | 2020-02-18 |
Family
ID=69488446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910975438.4A Pending CN110807315A (en) | 2019-10-15 | 2019-10-15 | Topic model-based online comment emotion mining method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110807315A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113934846A (en) * | 2021-10-18 | 2022-01-14 | 华中师范大学 | Online forum topic modeling method combining behavior-emotion-time sequence |
| CN114580431A (en) * | 2022-02-28 | 2022-06-03 | 山西大学 | Dynamic theme quality evaluation method based on optimal transportation |
| CN114792246A (en) * | 2022-03-02 | 2022-07-26 | 西安邮电大学 | Method and system for mining typical product characteristics based on topic integration clustering |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631018A (en) * | 2015-12-29 | 2016-06-01 | 上海交通大学 | Article feature extraction method based on topic model |
| CN109101478A (en) * | 2018-06-04 | 2018-12-28 | 东南大学 | A kind of Aspect grade sentiment analysis method towards electric business comment text |
| CN109933657A (en) * | 2019-03-21 | 2019-06-25 | 中山大学 | A kind of Topics Crawling sentiment analysis method based on user characteristics optimization |
| CN109977413A (en) * | 2019-03-29 | 2019-07-05 | 南京邮电大学 | A kind of sentiment analysis method based on improvement CNN-LDA |
-
2019
- 2019-10-15 CN CN201910975438.4A patent/CN110807315A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631018A (en) * | 2015-12-29 | 2016-06-01 | 上海交通大学 | Article feature extraction method based on topic model |
| CN109101478A (en) * | 2018-06-04 | 2018-12-28 | 东南大学 | A kind of Aspect grade sentiment analysis method towards electric business comment text |
| CN109933657A (en) * | 2019-03-21 | 2019-06-25 | 中山大学 | A kind of Topics Crawling sentiment analysis method based on user characteristics optimization |
| CN109977413A (en) * | 2019-03-29 | 2019-07-05 | 南京邮电大学 | A kind of sentiment analysis method based on improvement CNN-LDA |
Non-Patent Citations (2)
| Title |
|---|
| XIANGFENG LUO，YAWEN YI: "Topic-Specific Emotion Mining Model for Online Comments" * |
| 何祥，骆祥峰: "基于关联语义链网络的文本聚类方法" * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113934846A (en) * | 2021-10-18 | 2022-01-14 | 华中师范大学 | Online forum topic modeling method combining behavior-emotion-time sequence |
| CN114580431A (en) * | 2022-02-28 | 2022-06-03 | 山西大学 | Dynamic theme quality evaluation method based on optimal transportation |
| CN114792246A (en) * | 2022-03-02 | 2022-07-26 | 西安邮电大学 | Method and system for mining typical product characteristics based on topic integration clustering |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110807315A (en) | Topic model-based online comment emotion mining method | |
| Hamouda et al. | Building machine learning based senti-word lexicon for sentiment analysis | |
| EP3203383A1 (en) | Text generation system | |
| Ljajić et al. | Improving sentiment analysis for twitter data by handling negation rules in the Serbian language | |
| Ramalingam et al. | Emotion detection from text | |
| CN108388660A (en) | A kind of improved electric business product pain spot analysis method | |
| Cao et al. | Overview of text visualization techniques | |
| Pramodh et al. | Automatic personality recognition of authors using big five factor model | |
| CN108388554A (en) | Text emotion identifying system based on collaborative filtering attention mechanism | |
| CN106446147A (en) | Emotion analysis method based on structuring features | |
| Seale et al. | Keyword analysis: A new tool for qualitative research | |
| Cao et al. | Sentiment analysis based on expanded aspect and polarity-ambiguous word lexicon | |
| Rakhsha et al. | Detecting adverse drug reactions from social media based on multichannel convolutional neural networks modified by support vector machine | |
| CN110096597A (en) | A kind of text TF-IDF feature reconstruction method of combination emotional intensity | |
| CN111339403B (en) | Commodity comment-based new word extraction method | |
| CN110765762B (en) | System and method for extracting optimal theme of online comment text under big data background | |
| Sigit et al. | Comparison of Classification Methods on Sentiment Analysis of Political Figure Electability Based on Public Comments on Online News Media Sites | |
| Abdelrazaq et al. | A machine learning system for distinguishing nominal and verbal Arabic sentences | |
| Sani et al. | Sentiment Analysis of Hausa Language Tweet Using Machine Learning Approach | |
| Rarasati et al. | Correlation between Twitter sentiment analysis with three kernels using algorithm support vector machine (SVM) governor candidate electability level | |
| Roman et al. | Sentiment and behaviour annotation in a corpus of dialogue summaries | |
| Haney | Sentiment analysis: Providing categorical insight into unstructured textual data | |
| Vijayaragavan et al. | Online Prediction of Positive and Negative Emotionology Applying Machine Learning Technique | |
| Wang et al. | Research on Feature Mining Algorithm Based on Product Reviews | |
| Chauhan et al. | Sentiment Analysis and Predictions of COVID 19 Tweets using Natural Language Processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200218 |