CN113590744A - Interpretable emotion tracing method - Google Patents
Interpretable emotion tracing method Download PDFInfo
- Publication number
- CN113590744A CN113590744A CN202110743688.2A CN202110743688A CN113590744A CN 113590744 A CN113590744 A CN 113590744A CN 202110743688 A CN202110743688 A CN 202110743688A CN 113590744 A CN113590744 A CN 113590744A
- Authority
- CN
- China
- Prior art keywords
- emotion
- model
- result
- training
- rule
- 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 98
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000004458 analytical method Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims description 40
- 238000012360 testing method Methods 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 11
- 238000002372 labelling Methods 0.000 claims description 11
- 230000009193 crawling Effects 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 5
- 230000007935 neutral effect Effects 0.000 claims description 4
- 238000013461 design Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012821 model calculation Methods 0.000 claims description 2
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000011161 development Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002996 emotional effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012706 support-vector machine 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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/103—Formatting, i.e. changing of presentation of documents
- G06F40/117—Tagging; Marking up; Designating a block; Setting of attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/166—Editing, e.g. inserting or deleting
- G06F40/186—Templates
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Machine Translation (AREA)
Abstract
The invention provides an interpretable emotion tracing method, which provides an emotion tracing concept for the first time aiming at emotion analysis problems, wherein emotion tracing refers to the fact that a complete reasoning link is found in an emotion analysis process and the reasoning link has text support; the emotion analysis process is made to be more interpretable by utilizing the rule template; the text support is searched for the extracted rule by utilizing the traceability evidence model, and the text support is considered to be a correct rule only if the rule with the text support is included and added into the result, so that the result is more accurate; the traceability evidence model is verified to be beneficial to emotion traceability, so that emotion analysis is facilitated, and the method has good practicability.
Description
Technical Field
The invention relates to the field of natural language of computer technology, in particular to an interpretable emotion tracing method.
Background
Traceability reasoning has been studied in both formal logic and natural language processing. Traceability reasoning means that given a theory C and possibly a hint Q, this hint cannot be demonstrated in theory C. There is a new fact f, adding f to C, which can give an indication of Q, in NLP, f can be masked from C by masking.
Emotion analysis has received attention from many researchers in recent years due to the rapid development of social media, and with the rapid development of social media, information such as comments, forum discussions, and opinions of social platforms appears in a large amount, so that the information mining expression of emotions of entities such as things, events or topics is helpful for understanding the opinions of the people and the choices made by the people. Since the beginning of 2000, emotion analysis has become one of the most active research areas in natural language processing.
The emotion analysis is to carry out emotion classification on information, and some tasks are divided into two types: positive or negative, some tasks add neutral classification on this basis. Early emotion analysis methods included supervised methods based on traditional machine learning methods such as support vector machines, maximum entropy, naive bayes, etc., and unsupervised methods such as emotion dictionaries, etc. Later, as machine learning developed, the method of machine learning was also used in the task of emotion analysis, for example, Kalchbrenner et al proposed a dynamic CNN model (DCNN) that uses a dynamic K-Max pooling algorithm as a non-linear sub-sampling function, with feature maps generated by the network, able to capture relationships between words. With the development of emotion analysis, the t et al proposed the task of aspect emotion analysis, which requires consideration of both emotion and target information, the target usually being an entity or aspect, i.e. given a sentence and a target aspect, aspect level emotion classification aims to infer the emotional polarity of the sentence for the specified target aspect.
The existing method is basically an unsupervised emotion dictionary method or a supervised deep learning network method, but the existing model has the problems of poor interpretability and poor mobility, and for texts in another field, a great amount of marking and training may be required again by using the supervised method.
Disclosure of Invention
The invention provides an interpretable emotion tracing method, which solves the problem that an emotion analysis model in the past is not high in interpretability, and obtains a more convincing result through emotion tracing.
In order to achieve the technical effects, the technical scheme of the invention is as follows:
an interpretable emotion tracing method comprises the following steps:
s1: collecting data and preprocessing the collected data;
s2: extracting a model according to a training rule;
s3: training a traceability evidence model;
s4: forming an emotion traceability result;
s5: and inputting the unlabeled text to obtain an emotion traceability result.
Further, the specific process of step S1 is:
s11: the comment information is crawled from a social platform or a shopping platform;
s12: for the crawled information, cleaning and removing parts which do not relate to evaluation, and reserving information expressing emotion or opinion of the user on the commodity or the entity;
s13: marking the information obtained by crawling manually, wherein the marked content comprises: the method comprises the steps of obtaining the corresponding emotion polarity, obtaining the reasoning process of the result, templating the reasoning process to obtain a rule template, and marking which clauses belong to the evidence which needs to be searched in the text and which belong to the external knowledge in the rule template.
Further, the specific process of step S2 is:
s21: dividing a data set into a training set and a testing set, wherein the training set is 70 percent, and the testing set is 30 percent;
s22: the rule extraction model is a sequence labeling question-answer model, and a linear layer of sequence labeling is added on the basis of a pre-training language model Roberta model;
s23: model training covers the part of each labeled reasoning rule clause, which belongs to the template variable, converts the part into a question sentence, takes the labeled variable result as the answer of the question, inputs the question sentence and the text, and trains the model to enable the highest evaluation sequence to be the correct answer. All related clauses of all training sets are used as one-time complete training;
s24: in the training process, the model trained each time is recorded, and the model with the highest test set accuracy and single f1 value is obtained through the test set test.
Further, the specific process of step S3 is:
s31: the data set for this step was constructed: and replacing the same category in the sentence by the variable related in the rule template in a similar replacement mode to obtain a new inference clause. Marking the clauses generated in the mode as false, and marking the originally marked reasoning chain clause as true; the training set accounts for 70%, and the testing set accounts for 30%;
s32: the tracing evidence model is that sentences in a text are spliced with correspondingly coded clauses, the middle of the sentence is divided by using a special separator of Roberta, the sentence is input into a Roberta pre-training language model after being coded, vectors of the character are added, and the added vectors are input into a linear layer and softmax normalization to obtain the condition of judging whether corresponding variables meet the description condition of the clauses or not;
s33: in the training process, the model is stored once every time the training of all samples is completed, and finally the model with the highest accuracy of the test set is used.
Further, the specific process of step S4 is:
s41: obtaining the score of the whole emotion traceability link through the score of each clause obtained in the step S3;
s42: comparing the scores of all the links, and taking the link with the highest score in each aspect as a final emotion traceability result;
s43: after obtaining the emotion tracing result, obtaining the emotion analysis result of the session on the aspect through the conclusion of the emotion tracing reasoning chain.
Further, the specific process of step S5 is:
s51: for each rule template, splitting a clause, covering variables and constructing a problem text;
s52: inputting the questions and texts obtained in the last step into a rule extraction model to obtain answers to the questions, replacing variables, and constructing texts of the questions in the next step, wherein if the answers do not exist, the emotion tracing of the texts can be considered to be incapable of being carried out by using the template;
s53: repeating S51 and S52 until all variables of all templates have been confirmed, i.e., confirmed as specific content or confirmed as not found;
s54: inputting clauses related to the evidence needing to be searched from the text and sentences with corresponding variables mentioned in the text in the S53 into a traceability evidence model to obtain the scores of the clauses so as to obtain the scores of the results;
s54: and obtaining the score of the complete emotion traceability link from the score obtained in the last step, thereby obtaining a final result.
Further, in step S1, crawling comment information from a social network site or a shopping network site by using a crawler tool, only preserving the existence of comments expressing the emotion of the user, manually labeling the data with the polarity of the emotion of the user, wherein the polarity is positive, negative and neutral, adding the used rule, and meanwhile, the labeling also needs to extract the rule template from the labeled rule as an input; in step S2, the rule template is split and converted into question sentences, and a question-answer model is trained to obtain information of corresponding variables in the template; training a traceability evidence model according to the result obtained in the step S2, and scoring sentences related to sentence information rules in the step S2 by model calculation of all sentences in the text which are referred to related variables, so as to score the combined rules; and setting a threshold value for the score in the step S4 to obtain the score of each inference link, and determining a final emotion inference link to obtain a final result.
Further, in step S5, the unlabeled text is matched with the rule template, the variables in the rule template are covered to design a question sentence, the result of the variable is obtained through the rule extraction model, the result of the next variable is input into the template, the result of all the variables in the template sentence is known to be found, and a plurality of emotion traceability inference links are obtained. And obtaining the score of each link clause through a traceability evidence model according to the result obtained in the last step, using the highest score in the same aspect as an emotion traceability result, and using the link result as an emotion analysis conclusion.
Compared with the prior art, the technical scheme of the invention has the beneficial effects that:
aiming at the emotion analysis problem, the emotion tracing concept is firstly provided, and the emotion tracing refers to the fact that a complete reasoning link is found in the emotion analysis process and the reasoning link has text support; the emotion analysis process is made to be more interpretable by utilizing the rule template; the text support is searched for the extracted rule by utilizing the traceability evidence model, and the text support is considered to be a correct rule only if the rule with the text support is included and added into the result, so that the result is more accurate; the traceability evidence model is verified to be beneficial to emotion traceability, so that emotion analysis is facilitated, and the method has good practicability.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a flow chart of the method application phase of the present invention.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the patent;
for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product;
it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.
1-2, an interpretable sentiment tracing method includes the following steps:
step 1, preprocessing such as data crawling, cleaning and labeling: the method comprises the steps of crawling comment information from social websites, shopping websites and the like by using a crawler tool, cleaning some meaningless texts, only keeping comments expressing the emotion of a user, manually marking the data according to the emotion polarity of the user, and adding a using rule into the data, wherein the polarity is positive, negative and neutral. Meanwhile, labeling also requires extracting a rule template from the labeled rule as input:
s11: the comment information is crawled from a social platform or a shopping platform;
s12: for the crawled information, cleaning and removing parts which do not relate to evaluation, and reserving information expressing emotion or opinion of the user on the commodity or the entity;
s13: marking the information obtained by crawling manually, wherein the marked content comprises: the method comprises the steps of obtaining the corresponding emotion polarity, obtaining the reasoning process of the result, templating the reasoning process to obtain a rule template, and marking which clauses belong to the evidence which needs to be searched in the text and which belong to the external knowledge in the rule template.
Step 2: training an evidence extraction model: extracting relevant evidences by using an evidence extraction model through the paragraphs obtained in the step 2, wherein the evidences are expressed in a triple mode, and the relationship is defined in the process of constructing the data set:
s21: dividing a data set into a training set and a testing set, wherein the training set accounts for 70% and the testing set accounts for 30%;
s22: the rule extraction model is a sequence labeling question-answer model, and a linear layer of sequence labeling is added on the basis of a pre-training language model Roberta model;
s23: model training covers the part of each labeled reasoning rule clause, which belongs to the template variable, converts the part into a question sentence, takes the labeled variable result as the answer of the question, inputs the question sentence and the text, and trains the model to enable the highest evaluation sequence to be the correct answer. All related clauses of all training sets are used as one-time complete training;
s24: in the training process, recording the model trained each time, and obtaining the model with the highest test set accuracy and single f1 value through test set testing;
s25: in use, the answer obtained by the previous question is taken to the value of the variable and then taken to the next clause, constituting the details of the next clause. For example for a rule template: if A is positive, B is A, B belongs to C, then C is positive, the question of the first clause: what is positive? After answer a is obtained, let a equal the answer, replace to the second clause, the question sentence becomes: what is a? And so on until each variable of the entire rule template finds a specific answer.
And step 3: and (3) composing an emotion traceability result: setting a threshold value for the score in the step 4 to obtain the score of each inference link, determining a final emotion inference link, and obtaining a final result:
s31: the data set for this step was constructed: and replacing the same category in the sentence by the variable related in the rule template in a similar replacement mode to obtain a new inference clause. The clauses generated in this way are marked as false, and the originally marked inference chain clause is marked as true. The training set accounts for 70%, and the testing set accounts for 30%;
s32: the tracing evidence model is that sentences in a text are spliced with correspondingly coded clauses, the middle of the sentence is divided by using a special separator of Roberta, the sentence is input into a Roberta pre-training language model after being coded, vectors of the character are added, and the added vectors are input into a linear layer and softmax normalization to obtain the condition of judging whether corresponding variables meet the description condition of the clauses or not;
s33: in the training process, parameters are updated in a random gradient descent mode, a model is stored once each time training of all samples is completed, and finally the model with the highest test set accuracy is used.
And 4, step 4: and (3) composing an emotion traceability result: setting a threshold value for the score in the step 4 to obtain the score of each inference link, determining a final emotion inference link, and obtaining a final result:
s41: obtaining the score of the whole emotion traceability link through the score of each clause obtained in the step 3;
s42: comparing the scores of all the links, and taking the link with the highest score in each aspect as a final emotion traceability result;
s43: after obtaining the emotion tracing result, obtaining the emotion analysis result of the session on the aspect through the conclusion of the emotion tracing reasoning chain.
And 5: inputting an unlabeled text to obtain an emotion traceability result: the method comprises the steps of matching an unlabeled text with a rule template, covering variables in the rule template to design question sentences, extracting a result of the variable through a rule extraction model, inputting the result into the template, continuously obtaining a result of the next variable, knowing that all the variables in the template sentence have found the result, and obtaining a plurality of emotion traceability reasoning links. And (3) obtaining the score of each link clause through a traceability evidence model according to the result obtained in the last step, using the highest score in the same aspect as an emotion traceability result, and using the link result as the conclusion of emotion analysis:
s51: for each rule template, splitting a clause, covering variables and constructing a problem text;
s52: inputting the questions and texts obtained in the last step into a rule extraction model to obtain answers to the questions, replacing variables, and constructing texts of the questions in the next step, wherein if the answers do not exist, the emotion tracing of the texts can be considered to be incapable of being carried out by using the template;
s53: repeating S51 and S52 until all variables of all templates have been confirmed, i.e., confirmed as specific content or confirmed as not found;
s54: inputting clauses related to the evidence needing to be searched from the text and sentences with corresponding variables mentioned in the text in the S53 into a traceability evidence model to obtain the scores of the clauses so as to obtain the scores of the results;
s54: and obtaining the score of the complete emotion traceability link from the score obtained in the last step, thereby obtaining a final result.
The same or similar reference numerals correspond to the same or similar parts;
the positional relationships depicted in the drawings are for illustrative purposes only and are not to be construed as limiting the present patent;
it should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.
Claims (10)
1. An interpretable emotion tracing method is characterized by comprising the following steps:
s1: collecting data and preprocessing the collected data;
s2: extracting a model according to a training rule;
s3: training a traceability evidence model;
s4: forming an emotion traceability result;
s5: and inputting the unlabeled text to obtain an emotion traceability result.
2. The interpretable emotion tracing method according to claim 1, wherein the specific process of step S1 is:
s11: the comment information is crawled from a social platform or a shopping platform;
s12: for the crawled information, cleaning and removing parts which do not relate to evaluation, and reserving information expressing emotion or opinion of the user on the commodity or the entity;
s13: marking the information obtained by crawling manually, wherein the marked content comprises: the method comprises the steps of obtaining the corresponding emotion polarity, obtaining the reasoning process of the result, templating the reasoning process to obtain a rule template, and marking which clauses belong to the evidence which needs to be searched in the text and which belong to the external knowledge in the rule template.
3. The interpretable emotion tracing method according to claim 2, wherein the specific process of step S2 is:
s21: dividing a data set into a training set and a testing set, wherein the training set is 70 percent, and the testing set is 30 percent;
s22: the rule extraction model is a sequence labeling question-answer model, and a linear layer of sequence labeling is added on the basis of a pre-training language model Roberta model;
s23: model training covers the part of each labeled inference rule clause, which belongs to the template variable, and converts the part into a question clause, the labeled variable result is used as the answer of the question, the question clause and the text are input, the model is trained so that the highest evaluation sequence is the correct answer, and all related clauses of all training sets are used as one-time complete training;
s24: in the training process, the model trained each time is recorded, and the model with the highest test set accuracy and single f1 value is obtained through the test set test.
4. The interpretable emotion tracing method according to claim 3, wherein the specific process of step S3 is:
s31: the data set for this step was constructed: the method comprises the steps of replacing the same category in a sentence by the variable related in a rule template in a similar replacement mode to obtain a new reasoning clause, marking the clause generated in the mode as false, and marking the originally marked reasoning chain clause as true; the training set accounts for 70%, and the testing set accounts for 30%;
s32: the tracing evidence model is that sentences in a text are spliced with correspondingly coded clauses, the middle of the sentence is divided by using a special separator of Roberta, the sentence is input into a Roberta pre-training language model after being coded, vectors of the character are added, and the added vectors are input into a linear layer and softmax normalization to obtain the condition of judging whether corresponding variables meet the description condition of the clauses or not;
s33: in the training process, the model is stored once every time the training of all samples is completed, and finally the model with the highest accuracy of the test set is used.
5. The interpretable emotion tracing method according to claim 4, wherein the specific process of step S4 is:
s41: obtaining the score of the whole emotion traceability link through the score of each clause obtained in the step S3;
s42: comparing the scores of all the links, and taking the link with the highest score in each aspect as a final emotion traceability result;
s43: after obtaining the emotion tracing result, obtaining the emotion analysis result of the session on the aspect through the conclusion of the emotion tracing reasoning chain.
6. The interpretable emotion tracing method according to claim 5, wherein the specific process of step S5 is:
s51: for each rule template, splitting a clause, covering variables and constructing a problem text;
s52: inputting the questions and texts obtained in the last step into a rule extraction model to obtain answers to the questions, replacing variables, and constructing texts of the questions in the next step, wherein if the answers do not exist, the emotion tracing of the texts can be considered to be incapable of being carried out by using the template;
s53: repeating S51 and S52 until all variables of all templates have been confirmed, i.e., confirmed as specific content or confirmed as not found;
s54: inputting clauses related to the evidence needing to be searched from the text and sentences with corresponding variables mentioned in the text in the S53 into a traceability evidence model to obtain the scores of the clauses so as to obtain the scores of the results;
s54: and obtaining the score of the complete emotion traceability link from the score obtained in the last step, thereby obtaining a final result.
7. The interpretable emotion tracing method as claimed in claim 6, wherein in step S1, a crawler tool is used to crawl comment information from social websites and shopping websites, only comments expressing the emotion of the user are kept, the data is artificially labeled with the polarity of the emotion, the polarity is divided into positive, negative and neutral, and the used rules are added, and at the same time, the labeling also needs to extract rule templates from the labeled rules as input.
8. The interpretable emotion tracing method as claimed in claim 7, wherein in step S2, the question-answer model is trained by splitting the rule template and converting it into question sentences, so as to obtain information of corresponding variables in the template.
9. The interpretable emotion tracing method as claimed in claim 8, wherein a traceability evidence model is trained for the result obtained in step S2, all sentences in the text which refer to relevant variables are scored by the model calculation for sentence related information rules in step S2, so as to score the combined rules; and setting a threshold value for the score in the step S4 to obtain the score of each inference link, and determining a final emotion inference link to obtain a final result.
10. The interpretable emotion tracing method as claimed in claim 8, wherein in step S5, the unlabeled text is matched with the rule template, the variables in the rule template are covered to design an interrogative sentence, the result of the variables is obtained through the rule extraction model, the result of the next variable is input into the template, the result of the next variable is obtained continuously, it is known that all the variables in the template sentence have found the result, and a plurality of emotion tracing reasoning links are obtained; and obtaining the score of each link clause through a traceability evidence model according to the result obtained in the last step, using the highest score in the same aspect as an emotion traceability result, and using the link result as an emotion analysis conclusion.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110743688.2A CN113590744B (en) | 2021-06-30 | 2021-06-30 | Expandable emotion tracing method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110743688.2A CN113590744B (en) | 2021-06-30 | 2021-06-30 | Expandable emotion tracing method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113590744A true CN113590744A (en) | 2021-11-02 |
CN113590744B CN113590744B (en) | 2024-05-24 |
Family
ID=78245568
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110743688.2A Active CN113590744B (en) | 2021-06-30 | 2021-06-30 | Expandable emotion tracing method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113590744B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140365208A1 (en) * | 2013-06-05 | 2014-12-11 | Microsoft Corporation | Classification of affective states in social media |
US20170011029A1 (en) * | 2013-05-09 | 2017-01-12 | Moodwire, Inc. | Hybrid human machine learning system and method |
CN109101597A (en) * | 2018-07-31 | 2018-12-28 | 中电传媒股份有限公司 | A kind of electric power news data acquisition system |
CN110083700A (en) * | 2019-03-19 | 2019-08-02 | 北京中兴通网络科技股份有限公司 | A kind of enterprise's public sentiment sensibility classification method and system based on convolutional neural networks |
CN110427549A (en) * | 2019-06-20 | 2019-11-08 | 北京清博大数据科技有限公司 | A kind of network public opinion Source Tracing method, apparatus, terminal and storage medium |
CN110612525A (en) * | 2017-05-10 | 2019-12-24 | 甲骨文国际公司 | Enabling thesaurus analysis by using an alternating utterance tree |
CN110968699A (en) * | 2019-11-01 | 2020-04-07 | 数地科技(北京)有限公司 | Logic map construction and early warning method and device based on event recommendation |
CN112000804A (en) * | 2020-08-18 | 2020-11-27 | 安徽理工大学 | Microblog hot topic user group emotion tendentiousness analysis method |
-
2021
- 2021-06-30 CN CN202110743688.2A patent/CN113590744B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170011029A1 (en) * | 2013-05-09 | 2017-01-12 | Moodwire, Inc. | Hybrid human machine learning system and method |
US20140365208A1 (en) * | 2013-06-05 | 2014-12-11 | Microsoft Corporation | Classification of affective states in social media |
CN110612525A (en) * | 2017-05-10 | 2019-12-24 | 甲骨文国际公司 | Enabling thesaurus analysis by using an alternating utterance tree |
CN109101597A (en) * | 2018-07-31 | 2018-12-28 | 中电传媒股份有限公司 | A kind of electric power news data acquisition system |
CN110083700A (en) * | 2019-03-19 | 2019-08-02 | 北京中兴通网络科技股份有限公司 | A kind of enterprise's public sentiment sensibility classification method and system based on convolutional neural networks |
CN110427549A (en) * | 2019-06-20 | 2019-11-08 | 北京清博大数据科技有限公司 | A kind of network public opinion Source Tracing method, apparatus, terminal and storage medium |
CN110968699A (en) * | 2019-11-01 | 2020-04-07 | 数地科技(北京)有限公司 | Logic map construction and early warning method and device based on event recommendation |
CN112000804A (en) * | 2020-08-18 | 2020-11-27 | 安徽理工大学 | Microblog hot topic user group emotion tendentiousness analysis method |
Non-Patent Citations (1)
Title |
---|
陈国兰: "基于情感词典与语义规则的微博情感分析", 情报探索, no. 2, 29 February 2016 (2016-02-29), pages 1 - 6 * |
Also Published As
Publication number | Publication date |
---|---|
CN113590744B (en) | 2024-05-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Amin et al. | Bengali vader: A sentiment analysis approach using modified vader | |
CN110929034A (en) | Commodity comment fine-grained emotion classification method based on improved LSTM | |
CN108121702A (en) | Mathematics subjective item reads and appraises method and system | |
Zhao et al. | ZYJ123@ DravidianLangTech-EACL2021: Offensive language identification based on XLM-RoBERTa with DPCNN | |
CN110598219A (en) | Emotion analysis method for broad-bean-net movie comment | |
CN111309891B (en) | System for reading robot to automatically ask and answer questions and application method thereof | |
CN108052504A (en) | Mathematics subjective item answers the structure analysis method and system of result | |
CN112287106A (en) | Online comment emotion classification method based on dual-channel hybrid neural network | |
CN115906816A (en) | Text emotion analysis method of two-channel Attention model based on Bert | |
Karrabi et al. | Sentiment analysis of informal persian texts using embedding informal words and attention-based lstm network | |
Nandakumar et al. | Sentimental analysis on student feedback using NLP & POS tagging | |
CN113935459A (en) | Automatic scoring method of deep neural network model based on BERT | |
CN117592563A (en) | Power large model training and adjusting method with field knowledge enhancement | |
CN111815426A (en) | Data processing method and terminal related to financial investment and research | |
CN111414456A (en) | Method and system for automatically scoring open type short answer questions | |
CN114579706B (en) | Automatic subjective question review method based on BERT neural network and multi-task learning | |
Rahman et al. | A dynamic strategy for classifying sentiment from Bengali text by utilizing Word2vector model | |
CN113590744B (en) | Expandable emotion tracing method | |
Sajinika et al. | Twitter Sentiment Analysis and Topic Modeling for Online Learning | |
CN107992482A (en) | Mathematics subjective item answers the stipulations method and system of step | |
Faturohman et al. | Sentiment Analysis on Social Security Administrator for Health Using Recurrent Neural Network | |
Wu | English Composition Assisted Writing Training Based on Big Data Text Mining | |
CN110909736B (en) | Image description method based on long-term and short-term memory model and target detection algorithm | |
Meng et al. | Nonlinear network speech recognition structure in a deep learning algorithm | |
Güler et al. | A review on the science of 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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |