CN112966518B - High-quality answer identification method for large-scale online learning platform - Google Patents
High-quality answer identification method for large-scale online learning platform Download PDFInfo
- Publication number
- CN112966518B CN112966518B CN202011535456.XA CN202011535456A CN112966518B CN 112966518 B CN112966518 B CN 112966518B CN 202011535456 A CN202011535456 A CN 202011535456A CN 112966518 B CN112966518 B CN 112966518B
- Authority
- CN
- China
- Prior art keywords
- answer
- model
- answers
- comments
- questions
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 239000013598 vector Substances 0.000 claims abstract description 89
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000013145 classification model Methods 0.000 claims abstract description 16
- 238000002372 labelling Methods 0.000 claims abstract description 11
- 238000010276 construction Methods 0.000 claims abstract description 4
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 230000008451 emotion Effects 0.000 claims description 42
- 238000000605 extraction Methods 0.000 claims description 13
- 238000005516 engineering process Methods 0.000 claims description 7
- 230000007246 mechanism Effects 0.000 claims description 4
- 230000004931 aggregating effect Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000009193 crawling Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000013526 transfer learning Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 2
- 238000011156 evaluation Methods 0.000 abstract description 3
- 238000011160 research Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—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
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Educational Technology (AREA)
- Educational Administration (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- General Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Machine Translation (AREA)
Abstract
A high-quality answer identification method for a large-scale online learning platform comprises the following steps: (one), feature vector construction: after preprocessing the acquired data set, manually labeling the data set, and then constructing a feature vector; secondly, taking the feature vector constructed in the first step as input, taking the artificially marked label as output, constructing a XGBOOST-based classification model and training; thirdly, for a new question and a series of answers and comments thereof, constructing three feature vectors by using the text content of the new question, the text content of the answer and the text content of the comment, and inputting the three feature vectors into the trained model in the second step so as to obtain a series of classification results as a result of identifying high-quality answers; the invention uses more information with different angles, fully uses the questions, the answers and the answer comments to solve the problem of identifying high-quality answers, and improves the prediction result to a certain extent on a plurality of evaluation indexes.
Description
Technical Field
The invention relates to the technical field of artificial intelligence natural language processing, in particular to a high-quality answer identification method for a large-scale online learning platform.
Background
With the development of internet technology, online education is widely accepted by virtue of no time and place limitation, and more people learn by using an online learning mode, so that the online education is rapidly developed. Although the question-answering community provided by the large-scale online learning platform provides online communication opportunities for learners, because the number of learners is numerous, teachers cannot provide personalized and real-time question answering for students, so that the intelligent question-answering technology capable of simulating online of the teachers at any time becomes one of research hotspots of online education. How to quickly select the best answers to questions of learners becomes an important problem to be solved in the intelligent question-answering field.
The quality answer identification and answer sorting are essentially used for helping the user to obtain high-quality answers, so that the use experience of the user is improved. The difference between the two is that the answer ranking generally takes the praise number as a model learning target, but the praise number can only represent the answer quality to a certain extent, the praise number can be influenced by factors such as the publishing time of the answers, and the answer with the highest praise number does not represent the best answer. The answer ordering modes in the community question and answer platform mainly comprise the following steps: according to the content relativity, according to the length of the answer, according to the time of the answer, according to the quality answer, according to the comment number of the answer, according to the praise number of the answer, etc. The current large-scale online learning platform only provides a mode of sorting answers according to praise numbers and answer publishing time, and does not provide a function of identifying high-quality answers. For a large-scale online learning platform, providing an intelligent online question-answering service for simulating teachers at any time is an important way for improving user experience, and identifying high-quality answers is an important technology in intelligent question-answering.
At present, few studies on recognition of high-quality answers are relatively few, and answer ranking studies are most relevant to the study, and many researchers have proposed various answer ranking modes, as follows:
(1) A community question and answer platform answer ordering method (applicant: university of Chinese science and technology, application number: 201810186972.2);
(2) An answer ordering method for question and answer system (applicant: shenzhen research institute, university of Beijing, application number: 201810284245. X);
(3) Training method of answer quality determination model, and answer quality determination method and device (applicant: national information data Limited company, application number: 201811285467. X);
(4) A method for automatically identifying correct answers in a community question-answer forum based on artificial intelligence (applicant: beijing university of post, application number: 201911058818.8).
The related researches mainly take the praise number of the answers as a learning target of answer quality ranking, concentrate on evaluating the answer quality by using the characteristics of the relativity between the questions and the answers, the content attribute of the answers, the time attribute of the answers and the like, and neglect the positive influence of the comment text and the emotion polarity of the comment text of the answers on the answer quality evaluation.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a high-quality answer identification method for a large-scale online learning platform, which uses more information with different angles, fully uses questions, answers and answer comments to solve the problem of identifying high-quality answers, and improves the prediction result to a certain extent on a plurality of evaluation indexes.
In order to achieve the above object, the present invention is achieved by the following technical scheme.
A high-quality answer identification method for a large-scale online learning platform comprises the following steps:
construction of feature vectors: after preprocessing the acquired data set, manually labeling the data set, and then constructing the following three-angle feature vectors: semantic relevance features of questions and answers, document vector features of all comments of each answer, and emotion features of comments; the three angles of feature acquisition is realized in three ways: (1) Sentence vector representations of the questions and the answers are obtained, and then similarity of two semantic vectors is calculated based on cosine similarity, so that semantic relevance of the questions and the answers is obtained; (2) Using a HAN model to represent the document vector of the answer comment; (3) Extracting emotion characteristics of the comments by using transfer learning;
and (II) constructing a model: taking the feature vector constructed in the step (one) as input, taking the artificially marked label as output, constructing a XGBOOST-based classification model and training;
and thirdly, for a new question and a series of answers and comments thereof, constructing three feature vectors described in the first step by using the text content of the new question, the text content of the answer and the text content of the comments, and inputting the three feature vectors into the trained model in the second step so as to obtain a series of classification results as a result of identifying high-quality answers.
The manual labeling specific operation in the step (one) is as follows:
crawling website information by using a crawler technology, storing and sorting the information of questions, answers, answer comments and answer point praise numbers, clearing the data of the questions, the answers and the comments which are empty, integrating the comments of the same answer under the same question, storing the acquired data in the form of the questions, the answers and the integrated comments, and manually marking the crawled data set by using the following method:
in the above formula, flag represents the label of the text pair, if the answer is wrong, it is considered to be a poor answer, the text pair is marked as '0', if the answer is correct but imperfect, it is considered to be a normal answer, the text pair is marked as '1', if the answer is correct and perfect, it is considered to be a good answer, the text pair is marked as '2', and after the manual marking is completed, the final data set contains the following contents: labels of questions, answers, integrated answer comments and text pairs;
the semantic relevance feature extraction operation of the questions and answers is as follows:
(1) The method comprises the steps of obtaining sentence vectors of questions and answers by using a BERT model, inputting the texts of the questions and the answers into the BERT model, generating sentence vectors, and taking output values of a penultimate layer of the pre-training model as the sentence vectors of the questions and the answers;
(2) The cosine similarity method is used for calculating the similarity between the questions and the answers, and the cosine value of the included angle of the two vectors is calculated to measure the similarity between the questions and the answers.
The document vector feature extraction operation of the answer comments comprises the following steps:
extracting characteristics of a plurality of comments by using a hierarchical attention network HAN, wherein the HAN model is divided into two parts, one part is used for constructing sentence vectors according to word vectors, the other part is used for constructing document vectors according to sentence vectors, comment contents in a data set are used as input of the HAN model, labels of text pairs are used as output of the HAN model for model training, and the penultimate layer of the model is used as the document vectors of the comments;
the HAN model is a neural network for document classification, and has two features: firstly, with a hierarchical structure, a document vector can be constructed by first constructing a representation of a sentence and then aggregating it into a document representation; secondly, two levels of attention mechanisms are applied at word and sentence levels, so that the attention mechanisms can strengthen the representation of important content when constructing the document representation;
the extraction operation of the emotion characteristics of the answer comments comprises the following steps:
because the obtained answer comment content does not have related emotion labels, and the workload of manual labeling is very large, part of data is randomly labeled with emotion labels, and then a pseudo-label strategy in semi-supervised learning is adopted to solve the problem of insufficient training data: firstly, training the marked data by using an emotion classification model to obtain an optimal model, marking unmarked data by using the optimal model to carry out pseudo-tag marking, and then training all the data to improve the model effect, wherein the method specifically comprises the following steps:
(1) Training on the marked comment data, acquiring a comment sentence vector by using a BERT model, inputting a comment text into the BERT model, taking an output value of a penultimate layer of the pre-training model as a sentence vector of a question and an answer, reducing the dimension of the sentence vector by using a fully connected network, normalizing the sentence vector after the dimension reduction by using softmax, and using the result for emotion classification, thereby obtaining a trained emotion classification model; the emotion classification model consists of an input layer, a pre-trained BERT model, a fully-connected network layer and an output layer;
(2) Analyzing unlabeled comment texts by using the trained emotion classification model in the step (1), expressing the unlabeled comment texts into sentence vectors, and performing emotion feature analysis by using the trained model to obtain the emotion features of comments; and combining the original data with the existing labels and the data generated based on the pseudo tag strategy, and continuing training the emotion analysis model to obtain an optimal model.
The invention has the advantages that: the invention is oriented to the recognition of the high-quality answers of the online education platform, and performs feature extraction from three angles, namely the correlation features of the questions and the answers, the comment document vector features of the answers and the emotion features of the comments of the answers. Compared with other methods, the method uses more information from different angles, and the prediction result is improved to a certain extent on a plurality of evaluation indexes.
Drawings
FIG. 1 is a flow chart of an implementation of an embodiment of the present invention.
Fig. 2 is a model diagram of answer similarity to questions.
Fig. 3 is a model diagram of the HAN model.
Fig. 4 is a graph of an answer comment emotion feature extraction model.
Detailed Description
The invention will be described in further detail with reference to the drawings and detailed description.
Referring to fig. 1, a high-quality answer identification method for a large-scale online learning platform comprises the following steps:
construction of feature vectors: after preprocessing (including outlier deletion, format processing, etc.) the acquired dataset, the dataset is manually annotated, and then the following three angles of feature vectors are constructed: semantic relevance features of questions and answers, document vector features of all comments of each answer, and emotion features of comments; the three angles of feature acquisition is realized in three ways: (1) Sentence vector representations of the questions and the answers are obtained, and then similarity of two semantic vectors is calculated based on cosine similarity, so that semantic relevance of the questions and the answers is obtained; (2) Using a HAN model to represent the document vector of the answer comment; (3) Extracting emotion characteristics of the comments by using transfer learning;
and (II) constructing a model: taking the feature vector constructed in the step (one) as input, taking the artificially marked label as output, constructing a XGBOOST-based classification model and training;
and thirdly, for a new question and a series of answers and comments thereof, constructing three feature vectors described in the first step by using the text content of the new question, the text content of the answer and the text content of the comments, and inputting the three feature vectors into the trained model in the second step so as to obtain a series of classification results as a result of identifying high-quality answers.
The manual labeling specific operation in the step (one) is as follows:
the method comprises the steps of crawling website information by using a crawler technology, storing and sorting the information of questions, answers, answer comments and answer point praise numbers, clearing data of the questions, the answers and the comments which are empty, integrating the comments of the same answer under the same question, and storing the obtained data in the form of the questions, the answers and the integrated comments. Manually labeling the crawled data set by using the following method:
in the above formula, flag represents the label of the text pair, if the answer is wrong, it is considered to be a poor answer, the text pair is marked as '0', if the answer is correct but imperfect, it is considered to be a normal answer, the text pair is marked as '1', if the answer is correct and perfect, it is considered to be a good answer, and the text pair is marked as '2'. After manual annotation is completed, the final dataset contains the following: questions, answers, integrated answer comments, and labels for text pairs.
Referring to fig. 2, the semantic relevance feature extraction operation of the questions and answers is as follows:
(1) The method has the advantages that the BERT model is used for obtaining sentence vectors of questions and answers, the traditional word vector sentence vector generation mode has a large disadvantage, the same word can be expressed into the same vector when different context semantics are different, the BERT is a large pre-training model, the problem of word ambiguity can be solved, and good experimental results can be obtained by using the BERT and fine tuning in a specific field. The BERT includes two versions, a 12-layer transducer and a 24-layer transducer, the experiment uses a 12-layer model to perform the experiment, theoretically, the output value of each layer of transducer can be used as a sentence vector, and the best sentence vector should be known to take the penultimate layer by referring to experimental data, because the value of the last layer is too close to the target and the values of the previous layers are not sufficiently learned for the semantic information of the sentence. Inputting the text of the questions and the answers into the BERT model, generating sentence vectors, and taking the output value of the penultimate layer of the pre-training model as the sentence vectors of the questions and the answers.
(2) The cosine similarity method is used for calculating the similarity between the questions and the answers, and the cosine value of the included angle of the two vectors is calculated to measure the similarity between the questions and the answers.
Referring to fig. 3, the document vector feature extraction operation of the answer comments is as follows:
in general, there are several comments on an answer, and the existing work is divided into the following two types: one is to splice a plurality of comments to obtain a longer document, and then to extract the characteristics of the document; the other is to model each comment separately and then aggregate the modeled features. The invention does not need to distinguish between single comments, so the invention adopts a first mode to splice a plurality of comments into a document, and then uses a document vector feature extraction method to process the document, in particular to:
extracting characteristics of a plurality of comments by using a hierarchical attention network HAN, wherein the HAN model is divided into two parts, one part is used for constructing sentence vectors according to word vectors, the other part is used for constructing document vectors according to sentence vectors, comment contents in a data set are used as input of the HAN model, labels of text pairs are used as output of the HAN model for model training, and the penultimate layer of the model is used as the document vectors of the comments;
the HAN model is a neural network for document classification, and has two features: firstly, with a hierarchical structure, a document vector can be constructed by first constructing a representation of a sentence and then aggregating it into a document representation; secondly, two levels of attention mechanisms are applied at the word and sentence level, enabling it to strengthen the representation of important content when building a document representation.
Referring to fig. 4, the extraction operation of the emotion features of the answer comments is as follows:
because the obtained answer comment content does not have related emotion labels, and the workload of manual labeling is very large, part of data is randomly labeled with emotion labels, and then a pseudo-label strategy in semi-supervised learning is adopted to solve the problem of insufficient training data: firstly, training the marked data by using an emotion classification model to obtain an optimal model, marking unmarked data by using the optimal model to carry out pseudo-tag marking, and then training all the data to improve the model effect, wherein the method specifically comprises the following steps:
(1) Training on the existing marked comment corpus, using a BERT model to obtain a comment sentence vector, inputting comment texts into the BERT model, taking an output value of the penultimate layer of the pre-training model as a sentence vector of a question and an answer, using a fully connected network to reduce the dimension of the sentence vector, normalizing the dimension-reduced sentence vector by softmax, using the result for emotion classification, and simultaneously obtaining a trained emotion classification model; the emotion classification model consists of an input layer, a pre-trained BERT model, a fully-connected network layer and an output layer;
(2) Analyzing the unlabeled comment text by using the trained emotion classification model in the step (1), expressing the unlabeled comment text as sentence vectors, and performing emotion feature analysis by using the trained model to obtain the emotion features of comments.
In summary, based on the three feature extraction methods, the finally obtained feature vector format is [ similarity of answer to question, document vector of comment, emotion feature of comment ].
Claims (5)
1. A high-quality answer identification method for a large-scale online learning platform is characterized by comprising the following steps:
construction of feature vectors: after preprocessing the acquired data set, manually labeling the data set, and then constructing the following three-angle feature vectors: semantic relevance features of questions and answers, document vector features of all comments of each answer, and emotion features of comments; the three angles of feature acquisition is realized in three ways: (1) Sentence vector representations of the questions and the answers are obtained, and then similarity of two semantic vectors is calculated based on cosine similarity, so that semantic relevance of the questions and the answers is obtained; (2) Using a HAN model to represent the document vector of the answer comment; (3) Extracting emotion characteristics of the comments by using transfer learning;
and (II) constructing a model: taking the feature vector constructed in the step (one) as input, taking the artificially marked label as output, constructing a XGBOOST-based classification model and training;
and thirdly, for a new question and a series of answers and comments thereof, constructing three feature vectors described in the first step by using the text content of the new question, the text content of the answer and the text content of the comments, and inputting the three feature vectors into the trained model in the second step so as to obtain a series of classification results as a result of identifying high-quality answers.
2. The method for recognizing high-quality answers to a large-scale online learning platform of claim 1, wherein,
the manual labeling specific operation in the step (one) is as follows:
crawling website information by using a crawler technology, storing and sorting the questions, answers, answer comments and answer point approval information, clearing the data of the questions, the answers and the comments which are empty, integrating the comments of the same answer under the same question, and storing the acquired data in the form of the questions, the answers and the integrated comments; manually labeling the crawled data set by using the following method:
in the above formula, flag represents the label of the text pair, if the answer is wrong, it is considered to be a poor answer, the text pair is marked as '0', if the answer is correct but imperfect, it is considered to be a normal answer, the text pair is marked as '1', if the answer is correct and perfect, it is considered to be a good answer, the text pair is marked as '2', and after the manual marking is completed, the final data set contains the following contents: questions, answers, integrated answer comments, and labels for text pairs.
3. The method for recognizing high-quality answers to a large-scale online learning platform of claim 1, wherein,
the semantic relevance feature extraction operation of the questions and answers is as follows:
(1) The method comprises the steps of obtaining sentence vectors of questions and answers by using a BERT model, inputting the texts of the questions and the answers into the BERT model, generating sentence vectors, and taking output values of a penultimate layer of the pre-training model as the sentence vectors of the questions and the answers;
(2) The cosine similarity method is used for calculating the similarity between the questions and the answers, and the cosine value of the included angle of the two vectors is calculated to measure the similarity between the questions and the answers.
4. The method for recognizing high-quality answers to a large-scale online learning platform of claim 1, wherein,
the document vector feature extraction operation of the answer comments comprises the following steps:
extracting characteristics of a plurality of comments by using a hierarchical attention network HAN, wherein the HAN model is divided into two parts, one part is used for constructing sentence vectors according to word vectors, the other part is used for constructing document vectors according to sentence vectors, comment contents in a data set are used as input of the HAN model, labels of text pairs are used as output of the HAN model for model training, and the penultimate layer of the model is used as the document vectors of the comments;
the HAN model is a neural network for document classification, and has two features: firstly, with a hierarchical structure, a document vector can be constructed by first constructing a representation of a sentence and then aggregating it into a document representation; secondly, two levels of attention mechanisms are applied at the word and sentence level, enabling it to strengthen the representation of important content when building a document representation.
5. The method for recognizing high-quality answers to a large-scale online learning platform of claim 1, wherein,
the extraction operation of the emotion characteristics of the answer comments comprises the following steps:
because the obtained answer comment content does not have related emotion labels, and the workload of manual labeling is very large, part of data is randomly labeled with emotion labels, and then a pseudo-label strategy in semi-supervised learning is adopted to solve the problem of insufficient training data: firstly, training the marked data by using an emotion classification model to obtain an optimal model, marking unmarked data by using the optimal model to carry out pseudo-tag marking, and then training all the data to improve the model effect, wherein the method specifically comprises the following steps:
(1) Training on the marked comment data, acquiring a comment sentence vector by using a BERT model, inputting a comment text into the BERT model, taking an output value of a penultimate layer of the pre-training model as a sentence vector of a question and an answer, reducing the dimension of the sentence vector by using a fully connected network, normalizing the sentence vector after the dimension reduction by using softmax, and using the result for emotion classification, thereby obtaining a trained emotion classification model; the emotion classification model consists of an input layer, a pre-trained BERT model, a fully-connected network layer and an output layer;
(2) Analyzing unlabeled comment texts by using the trained emotion classification model in the step (1), expressing the unlabeled comment texts into sentence vectors, and performing emotion feature analysis by using the trained model to obtain the emotion features of comments; and combining the original data with the existing labels and the data generated based on the pseudo tag strategy, and continuing training the emotion analysis model to obtain an optimal model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011535456.XA CN112966518B (en) | 2020-12-22 | 2020-12-22 | High-quality answer identification method for large-scale online learning platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011535456.XA CN112966518B (en) | 2020-12-22 | 2020-12-22 | High-quality answer identification method for large-scale online learning platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112966518A CN112966518A (en) | 2021-06-15 |
| CN112966518B true CN112966518B (en) | 2023-12-19 |
Family
ID=76271262
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011535456.XA Active CN112966518B (en) | 2020-12-22 | 2020-12-22 | High-quality answer identification method for large-scale online learning platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112966518B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113486177A (en) * | 2021-07-12 | 2021-10-08 | 贵州电网有限责任公司 | Electric power field table column labeling method based on text classification |
| CN114444481B (en) * | 2022-01-27 | 2023-04-07 | 四川大学 | Sentiment analysis and generation method of news comment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012039686A1 (en) * | 2010-09-24 | 2012-03-29 | National University Of Singapore | Methods and systems for automated text correction |
| LU101290B1 (en) * | 2018-08-17 | 2019-11-29 | Univ Qilu Technology | Method, System, Storage Medium and Electric Device of Medical Automatic Question Answering |
| CN111259127A (en) * | 2020-01-15 | 2020-06-09 | 浙江大学 | Long text answer selection method based on transfer learning sentence vector |
| CN112069302A (en) * | 2020-09-15 | 2020-12-11 | 腾讯科技（深圳）有限公司 | Training method of conversation intention recognition model, conversation intention recognition method and device |
-
2020
- 2020-12-22 CN CN202011535456.XA patent/CN112966518B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012039686A1 (en) * | 2010-09-24 | 2012-03-29 | National University Of Singapore | Methods and systems for automated text correction |
| LU101290B1 (en) * | 2018-08-17 | 2019-11-29 | Univ Qilu Technology | Method, System, Storage Medium and Electric Device of Medical Automatic Question Answering |
| CN111259127A (en) * | 2020-01-15 | 2020-06-09 | 浙江大学 | Long text answer selection method based on transfer learning sentence vector |
| CN112069302A (en) * | 2020-09-15 | 2020-12-11 | 腾讯科技（深圳）有限公司 | Training method of conversation intention recognition model, conversation intention recognition method and device |
Non-Patent Citations (3)
| Title |
|---|
| 中文问答社区答案质量的评价研究:以知乎为例;王伟;冀宇强;王洪伟;郑丽娟;;图书情报工作(第22期);全文 * |
| 基于预训练模型的文本分类网络TextCGA;杨玮祺;杜晔;;现代计算机(第12期);全文 * |
| 附加情感特征的在线问答社区信息质量自动化评价;姜雯;许鑫;武高峰;;图书情报工作(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112966518A (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021031480A1 (en) | Text generation method and device | |
| CN109271506A (en) | A kind of construction method of the field of power communication knowledge mapping question answering system based on deep learning | |
| CN110737763A (en) | Chinese intelligent question-answering system and method integrating knowledge map and deep learning | |
| CN112115238A (en) | Question-answering method and system based on BERT and knowledge base | |
| CN110825867B (en) | Similar text recommendation method and device, electronic equipment and storage medium | |
| CN111985239A (en) | Entity identification method and device, electronic equipment and storage medium | |
| CN108804654A (en) | A kind of collaborative virtual learning environment construction method based on intelligent answer | |
| Cai et al. | Intelligent question answering in restricted domains using deep learning and question pair matching | |
| CN110287298A (en) | A kind of automatic question answering answer selection method based on question sentence theme | |
| CN113011196B (en) | Concept-enhanced representation and one-way attention-containing subjective question automatic scoring neural network model | |
| CN110348024A (en) | Intelligent identifying system based on legal knowledge map | |
| CN113962219A (en) | Semantic matching method and system for knowledge retrieval and question answering of power transformer | |
| CN111143672A (en) | Expert specialty scholars recommendation method based on knowledge graph | |
| CN110968708A (en) | Method and system for labeling education information resource attributes | |
| CN111368058A (en) | Question-answer matching method based on transfer learning | |
| CN112966518B (en) | High-quality answer identification method for large-scale online learning platform | |
| CN111552773A (en) | Method and system for searching key sentence of question or not in reading and understanding task | |
| CN111581364A (en) | Chinese intelligent question-answer short text similarity calculation method oriented to medical field | |
| CN112541347A (en) | Machine reading understanding method based on pre-training model | |
| CN114020871B (en) | Multi-mode social media emotion analysis method based on feature fusion | |
| CN112989811B (en) | History book reading auxiliary system based on BiLSTM-CRF and control method thereof | |
| CN115311465A (en) | Image description method based on double attention models | |
| Xia et al. | Question-answering using keyword entries in the Oil&Gas domain | |
| CN113901224A (en) | Knowledge distillation-based secret-related text recognition model training method, system and device | |
| CN112231449A (en) | Vertical field entity chain finger system based on multi-path recall |
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 |