CN108804654A - A kind of collaborative virtual learning environment construction method based on intelligent answer - Google Patents
A kind of collaborative virtual learning environment construction method based on intelligent answer Download PDFInfo
- Publication number
- CN108804654A CN108804654A CN201810581196.6A CN201810581196A CN108804654A CN 108804654 A CN108804654 A CN 108804654A CN 201810581196 A CN201810581196 A CN 201810581196A CN 108804654 A CN108804654 A CN 108804654A
- Authority
- CN
- China
- Prior art keywords
- answer
- question
- learning environment
- intelligent
- collaborative virtual
- 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
Abstract
The present invention relates to a kind of collaborative virtual learning environment construction method based on intelligent answer belongs to human-computer interaction technique field, including 1. acquire and build specific area teaching question and answer data set, to the question and answer of acquisition to pre-processing；2. building BI-LSTM-CRF models to question and answer to carrying out semantic analysis and feature extraction, the participle and sequence labelling of sentence are realized；3. building the answer confidence calculations model based on LSTM networks, input question and answer weigh the matching degree of question and answer pair by model to feature vector, and the answer feature vector of highest scoring is converted output；4. building collaborative virtual learning environment using the virtual reality technology based on Unity3D, intelligent answer engine is built, introduces trained deep learning Question-Answering Model, realizes the collaborative virtual learning environment structure based on intelligent answer.The present invention realizes the intelligent answer of specific area semantic understanding level using deep learning, and is applied among the collaborative virtual learning environment of teaching-oriented, and idea and method has been expanded for virtual classroom and human-computer interaction research.
Description
Technical field
The invention belongs to human-computer interaction technique fields, are related to a kind of collaborative virtual learning environment structure side based on intelligent answer
Method.
Background technology
Collaborative virtual learning environment is classroom instruction and the combination of virtual reality technology, mainly applies virtual reality skill
Art builds teaching scene and the content of courses in natural three-dimensional virtual environment true to nature, and specific teaching, manoeuvre behaviour are provided for student
Work and key to difficulty.The structure of collaborative virtual learning environment is related to virtual reality, human-computer interaction, multimedia technology, computer network
How equal aspects of contents are designed and are appreciated that user view and the collaborative virtual learning environment interacted with real-time intelligent are to need instantly
The major issue to be considered.Intelligent answer is an important research direction of natural language processing field, with traditional search engines
Difference, for question answering system as the another way for meeting information requirement, it not only allows for user to be putd question in the form of natural language, also
Can be that user directly returns to required accurate answer.From the user point of view, intelligent answer is that one kind being concisely and efficiently information and obtains
Take mode, user that can effectively evade the mistake of traditional search engines without finding answer from dense interminable list of relevant documents
Load problem improves Information Retrieval Efficiency.Current internet+epoch, intelligent answer can with flexible Application to such as intelligent customer service,
All various aspects of the daily lifes such as medical advice and chat robots open up more easily life style for people.
Traditional question and answer technology mainly has based on the common knowledge library, is based on web retrieval and is based on keyword retrieval, they
Core is keyword discriminance analysis, problem matching and candidate answers sequence.Since Chinese expression way is flexible, there is identical semanteme
The position that occurs of sentence its keyword it is indefinite, keyword Match in sequence, which tends not to meet retrieval, to be required, while being deposited in Chinese
In a large amount of synonym, entirely different keyword may contain identical semanteme, these tradition sides in problem and answer
Method is general lack of the understanding analysis to question and answer statement semantics level, and there is manually mark dependence is strong, accuracy rate is low, can expand
The drawbacks such as malleability difference.Instantly the domestic ripe Chinese intelligent Answer System towards specific area is few, the application of question answering system
Development is also accordingly received and is kept in check, the main reason is that at present the country increase income can be for reference standard Chinese question and answer data set it is opposite
Less, the standard question and answer knowledge base for building certain scale needs to expend comparable human and material resources.
Existing a part of patent, which puts forth an idea, constructs intelligent Answer System, and applies it to disappearing based on virtual environment
In the fields such as anti-, medical treatment, household, new application model is opened up for intelligent answer.Wherein, by intelligent answer and collaborative virtual learning environment
The idea of connected applications is also gradually brought into schedule, when student does experiment in collaborative virtual learning environment and encounters query, intelligent answer
System can provide the resolving ideas and method of problem for student, while the detailed data of student question also can be automatic by backstage
It compiles, further improves question answering system.The two favorably combines on escape conventional teaching time, space and teaching resource
Limitation, allow student collaborative virtual learning environment carry out immersion it is real-time observe, interact, test, answer questions etc. operations become can
Can, this situated learning process improves the academic knowledge of student to a deeper level on the basis of saving qualified teachers' material resource cost
And Thinking Skills.
In conclusion although intelligent answer achieves a degree of achievement in research in natural language processing field, such as
The Chinese standard question and answer data set of what structure specific area is still that current intelligent answer studies generally existing with training pattern
How short slab rationally utilizes deep learning to optimize semantic feature extraction process from semantic level anolytic sentence to a deeper level, such as
There is still a need for continue practical exploration for key issues of what opens up question answering system development new direction in conjunction with the prior art.
Invention content
In view of this, the purpose of the present invention is to provide a kind of using the high void based on intelligent answer of original, accuracy
Quasi- academic environment construction method.
In order to achieve the above objectives, the present invention provides the following technical solutions：
A kind of collaborative virtual learning environment construction method based on intelligent answer, includes the following steps：
S1：It acquires and builds specific area teaching question and answer data set, to the question and answer of acquisition to pre-processing；
S2：Build two-way length in short-term memory stick part random field BI-LSTM-CRF models to question and answer to carry out semantic analysis and
The participle and sequence labelling of sentence are realized in feature extraction；
S3：Build the answer confidence calculations model based on long short-term memory LSTM networks, input question and answer to feature vector,
The answer feature vector of highest scoring is converted output by the matching degree that question and answer pair are weighed by model；
S4：Collaborative virtual learning environment is built using the virtual reality technology based on Unity3D, intelligent answer engine is built, draws
Enter trained deep learning Question-Answering Model, completes the collaborative virtual learning environment structure based on intelligent answer.
Further, in step sl, question and answer data are obtained by web crawlers technology and the approach that artificially collects, structure is specific
Field teaching question and answer data set.
Further, in step sl, it is to the question and answer of collection first to carrying out duplicate removal, going empty behaviour to the pretreatment of question and answer pair
Make to reject the unanswered data of mess code, it is final to obtain standard question and answer data, calculate sentence using shallow-layer neural network Word2Vec
The word sequence conditional probability of son, is embedded into and term vector model is trained and is calculated.
Further, in step s 2, question and answer are included the following steps to carrying out semantic analysis and feature extraction：
S21：The standard question and answer data and term vector model obtained using step S1 are trained BI-LSTM-CRF models
Test；
S22：By text sequence X=(x the problem of vectorization1,x2,...,xn) it is input to trained network model,
The relevant information of the front and back word of BILSTM synthesis, CRF consider the context between output label, after model participle mark,
The corresponding annotated sequences of text sequence X (also known as characteristic sequence) y=(y1,y2,...,yn) predict that output is：
Wherein, A is state-transition matrix, and P is the output matrix of BILSTM networks, Ai,jIt indicates in sequential from i-th of state
It is transferred to j-th of shape probability of state, Pi,jIndicate that i-th of word is j-th of probability marked in input observation sequence.
Further, in step s3, candidate answers are carried out by the answer confidence calculations model based on LSTM networks
Confidence level sorts, while deleting the lower answer of confidence level, selects optimal answer；
Using reconcile cosine similarity and Euclidean distance confidence calculations method computational problem answer matching degree,
The confidence level cosine of problem answers and reconcile cosine similarity and Euclidean distance CosEuclid is respectively：
Scorecosine(VQ,VA(the V of)=0.5Q,VA)+0.5 (3)
Wherein, VQAnd VAThe text vector sequence of correspondence problem answer respectively, and Scorecosine(VQ,VA(the V of)=0.5Q,
VA)+0.5 it is that cosine similarity is normalized into [0,1] section.
Further, in step s 4, flat in Unity3D using virtual reality technology according to specific field research situation
Virtual classroom or laboratory academic environment are built on platform；Virtual teacher and the student of human-computer interaction can be carried out by creating, virtual
In academic environment corresponding question and answer shell script and function are configured for different role；
On the basis of step S1-S3, Unity3D platforms are based on, the model constructed is asked by the intelligence of structure
It answers engine to dock with corresponding specific virtual teaching environment, realizes the collaborative virtual learning environment structure based on intelligent answer.
Compared with prior art, it advantages of the present invention and has the beneficial effect that：
1. building specific teaching field Chinese question and answer number by various ways such as website reptile, file retrieval and artificial acquisitions
According to collection, traditional conventional data collection is compared, the data set is more with practical value in practical applications.
2. by combining the two-way LSTM networks in CRF and deep learning, realize to the deeper language of question and answer sentence
Reason and good sense solution and feature extraction, the model help to more fully understand sentence meaning and carry out participle and sequence labelling, are that answer is defeated
Go out accuracy rate and does a degree of contribution.
3. propose the answer confidence level sequencing selection method for being combined LSTM deep neural networks with CosEuclid, it should
Method parses the relationship between problem answers from semantic level, and problem and the parameter of answer are shared.It is answered compared to tradition
The Keywords matching pattern of case selection, the method can not only make up the Chinese expression elusive short slab of polysemy, Er Qie
Effect is more excellent in terms of the accuracy of flexibility, search efficiency and inquiry.
4. the Chinese intelligent Answer System of specific area is combined by proposition with collaborative virtual learning environment, classroom has not only been simplified
The process of education enables students not constrain place and is exchanged with teacher classmate progress question and answer, and student need not be completely dependent on and turn over
The timely answer that weight difficult point is felt uncertain can just be obtained by readding a large amount of classroom notes.The method widened the application field of question answering system with
Development volue enriches the content of collaborative virtual learning environment to a certain degree, enhances the practical value of collaborative virtual learning environment, is more preferable
Ground human-computer interaction provides thinking.
Description of the drawings
In order to keep the purpose of the present invention, technical solution and advantageous effect clearer, the present invention provides following attached drawing and carries out
Explanation：
Fig. 1 is the collaborative virtual learning environment construction method system framework figure provided by the invention based on intelligent answer；
Fig. 2 is the answer preference pattern frame diagram provided by the invention based on LSTM networks.
Specific implementation mode
Below in conjunction with attached drawing, the preferred embodiment of the present invention is described in detail.
Fig. 1 is the collaborative virtual learning environment construction method system framework figure proposed by the invention based on intelligent answer.It is a kind of
Based on intelligent answer collaborative virtual learning environment construction method, main task is structure specific area question and answer data set；Utilize depth
Learning model understands question and answer to statement semantics, and carries out associated confidence to problem answers and calculate sequence；The algorithm knot that will be obtained
Fruit drives the decision-making module of question answering system under collaborative virtual learning environment by intelligent answer engine, is taught by deep learning algorithm fusion
The lecture contents of teacher are identified the enquirement of collaborative virtual learning environment Students ' and make corresponding answer in real time, realized and asked based on intelligence
The collaborative virtual learning environment structure answered.
The collaborative virtual learning environment construction method based on intelligent answer described in the present embodiment includes the following steps：
100：It acquires and builds specific area teaching question and answer data set, to the question and answer of acquisition to pre-processing；
200：Build two-way length in short-term memory stick part random field (BI-LSTM-CRF) model to question and answer to carry out semantic analysis
And feature extraction, realize the participle and sequence labelling of sentence；
300：Structure based on long short-term memory (LSTM) network answer confidence calculations model, input question and answer to feature to
Amount weighs the matching degree of question and answer pair by model, and the answer feature vector of highest scoring is converted output；
400：Collaborative virtual learning environment is built using the virtual reality technology based on Unity3D；
500：It designs intelligent answer engine and introduces trained deep learning Question-Answering Model, realize the void based on intelligent answer
Quasi- academic environment structure.
Optionally, in step 100, can also include the steps of：
101：By artificially collecting and the technical limit spacings question and answer data such as web crawlers, structure specific area teaching question and answer data
Collection.One problem generally corresponds to 3-5 answer, includes an optimal answer in each problem answers list；
102：The question and answer of acquisition are pre-processed to data set, to question and answer to the operation such as carrying out duplicate removal compound word, removing stop words to pick
It is final to obtain standard question and answer data except the unanswered data of mess code；
103：It is 200 dimensions that term vector, which is arranged, to be embedded in dimension, and the word order of sentence is calculated using shallow-layer neural network Word2Vec
Row conditional probability is embedded into and term vector model is trained and is calculated.
Optionally, in step 200, can also include the steps of：
Step 201：The standard question and answer data and step 103 obtained using step 102 obtain term vector model to BI-
LSTM-CRF models are trained；
Step 202：The dimension of setting input word insertion vector is respectively 50,100,150,200 dimensions in a model,
Dropout percentages are set as 25% and 50%, show that the word insertion dimension that optimal sentence ingredient extracts is by contrast test
200 dimensions, dropout percentages are 25%；
Step 203：According to the test result of step 201 by the text sequence X=(x of vectorization1,x2,...,xn) be input to
Trained network model.BILSTM can integrate the relevant information of front and back word, and CRF is contemplated that front and back between output label
Relationship.Text sequence X is after model participle mark processing, corresponding annotated sequence (also known as feature vector) y=(y1,
y2,...,yn) predict that output is：
Wherein, A is state-transition matrix, and P is the output matrix of BILSTM networks, Ai,jIt indicates in sequential from i-th of state
It is transferred to j-th of shape probability of state, Pi,jIndicate that i-th of word is j-th of probability marked in input observation sequence.
Optionally, in step 300, can also include the steps of：
Step 301：The answer preference pattern based on LSTM networks is built, model framework is as shown in Figure 2.In model problem and
The parameter sharing of answer, model can sort to the confidence level of candidate answers, while delete the lower answer of confidence level, obtain optimal
It answers；
Step 302：By contrast experiment, setting LSTM answers select the input dimension of problem answers term vector in network for
100 dimensions, dropout percentages are 40%；
Step 303：Use the confidence calculations method computational problem answer of reconciliation cosine similarity and Euclidean distance
Relevant matches degree, the confidence level cosine of problem answers and reconcile CosEuclid points of cosine similarity and Euclidean distance
It is not：
Scorecosine(VQ,VA(the V of)=0.5Q,VA)+0.5 (3)
Wherein, VQAnd VAThe text vector sequence of correspondence problem answer respectively, and Scorecosine(VQ,VA(the V of)=0.5Q,
VA)+0.5 it is that cosine similarity is normalized into [0,1] section.
Optionally, in step 400, can also include the steps of：
Step 401：In conjunction with specific field research situation, void is built on Unity3D platforms using virtual reality technology
Quasi- classroom or laboratory academic environment；
Step 402：Virtual teacher and the student of human-computer interaction can be carried out by creating, and be different role in collaborative virtual learning environment
Configure corresponding question and answer shell script and function；
Optionally, in step 500, can also include the steps of：
Step 501：Based on Unity3D platform construction intelligent answer engines, the model constructed is drawn by intelligent answer
It holds up and is docked with corresponding specific virtual teaching environment.Step 3 and 4 arithmetic result driving question and answer system under collaborative virtual learning environment
The decision-making module of system identifies collaborative virtual learning environment Students ' by the teaching plan of giving lessons of deep learning algorithm fusion teacher
Corresponding answer in real time is putd question to and made, promotes students ' interest of study and power, realizes the Virtual Learning ring based on intelligent answer
Border is built.
Finally illustrate, preferred embodiment above is merely illustrative of the technical solution of the present invention and unrestricted, although logical
It crosses above preferred embodiment the present invention is described in detail, however, those skilled in the art should understand that, can be
Various changes are made to it in form and in details, without departing from claims of the present invention limited range.
Claims (6)
1. a kind of collaborative virtual learning environment construction method based on intelligent answer, it is characterised in that：Include the following steps：
S1：It acquires and builds specific area teaching question and answer data set, to the question and answer of acquisition to pre-processing；
S2：Build two-way length in short-term memory stick part random field BI-LSTM-CRF models to question and answer to carrying out semantic analysis and feature
Extraction, realizes the participle and sequence labelling of sentence；
S3：The answer confidence calculations model based on long short-term memory LSTM networks is built, input question and answer pass through feature vector
Model weighs the matching degree of question and answer pair, and the answer feature vector of highest scoring is converted output；
S4：Collaborative virtual learning environment is built using the virtual reality technology based on Unity3D, builds intelligent answer engine, introduces instruction
The deep learning Question-Answering Model perfected completes the collaborative virtual learning environment structure based on intelligent answer.
2. a kind of collaborative virtual learning environment construction method based on intelligent answer according to claim 1, it is characterised in that：?
In the step S1, by web crawlers technology and approach acquisition question and answer data are artificially collected, build the specific area teaching
Question and answer data set.
3. a kind of collaborative virtual learning environment construction method based on intelligent answer according to claim 1, it is characterised in that：?
In the step S1, the pretreatment to the question and answer pair is to the question and answer of collection first to carrying out duplicate removal, going do-nothing operation to reject
The unanswered data of mess code, it is final to obtain standard question and answer data, the word order of sentence is calculated using shallow-layer neural network Word2Vec
Row conditional probability is embedded into and term vector model is trained and is calculated.
4. a kind of collaborative virtual learning environment construction method based on intelligent answer according to claim 1, it is characterised in that：?
In the step S2, to the question and answer to carrying out semantic analysis and feature extraction, include the following steps：
S21：BI-LSTM-CRF models are trained using the obtained standard question and answer data of the step S1 and term vector model
Test；
S22：By text sequence X=(x the problem of vectorization1,x2,...,xn) it is input to trained network model, BILSTM
The relevant information of comprehensive front and back word, CRF consider the context between output label, after model participle mark, text sequence
Arrange the corresponding annotated sequence y=(y of X1,y2,...,yn) predict that output is：
Wherein, A is state-transition matrix, and P is the output matrix of BILSTM networks, Ai,jIt indicates to shift from i-th of state in sequential
To j-th of shape probability of state, Pi,jIndicate that i-th of word is j-th of probability marked in input observation sequence.
5. a kind of collaborative virtual learning environment construction method based on intelligent answer according to claim 1, it is characterised in that：?
In the step S3, confidence level sequence is carried out to candidate answers by the answer confidence calculations model based on LSTM networks, together
When delete the lower answer of confidence level, select optimal answer；
Use the matching degree of the confidence calculations method computational problem answer of reconciliation cosine similarity and Euclidean distance, problem
The confidence level cos ine of answer and reconcile cosine similarity and Euclidean distance CosEuclid is respectively：
Scorecosine(VQ,VA(the V of)=0.5Q,VA)+0.5 (3)
Wherein, VQAnd VAThe text vector sequence of correspondence problem answer respectively, and
Scorecosine(VQ,VA(the V of)=0.5Q,VA)+0.5 it is that cosine similarity is normalized into [0,1] section.
6. a kind of collaborative virtual learning environment construction method based on intelligent answer according to claim 1, it is characterised in that：?
In the step S4, according to specific field research situation, built on Unity3D platforms using virtual reality technology virtual
Classroom or laboratory academic environment；Virtual teacher and the student of human-computer interaction can be carried out by creating, for not in collaborative virtual learning environment
Corresponding question and answer shell script and function are configured with role；
On the basis of step S1-S3, Unity3D platforms are based on, the model constructed is drawn by the intelligent answer of structure
It holds up and is docked with corresponding specific virtual teaching environment, complete the collaborative virtual learning environment structure based on intelligent answer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810581196.6A CN108804654A (en) | 2018-06-07 | 2018-06-07 | A kind of collaborative virtual learning environment construction method based on intelligent answer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810581196.6A CN108804654A (en) | 2018-06-07 | 2018-06-07 | A kind of collaborative virtual learning environment construction method based on intelligent answer |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108804654A true CN108804654A (en) | 2018-11-13 |
Family
ID=64087686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810581196.6A Pending CN108804654A (en) | 2018-06-07 | 2018-06-07 | A kind of collaborative virtual learning environment construction method based on intelligent answer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108804654A (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109408627A (en) * | 2018-11-15 | 2019-03-01 | 众安信息技术服务有限公司 | A kind of answering method and system merging convolutional neural networks and Recognition with Recurrent Neural Network |
| CN109671320A (en) * | 2018-12-12 | 2019-04-23 | 广东小天才科技有限公司 | It is a kind of that exercising method and electronic equipment are calculated quickly based on interactive voice |
| CN109739958A (en) * | 2018-11-22 | 2019-05-10 | 普强信息技术（北京）有限公司 | A kind of specification handbook answering method and system |
| CN109766427A (en) * | 2019-01-15 | 2019-05-17 | 重庆邮电大学 | A kind of collaborative virtual learning environment intelligent answer method based on stacking Bi-LSTM network and collaboration attention |
| CN109885661A (en) * | 2019-02-27 | 2019-06-14 | 上海优谦智能科技有限公司 | Educate the question answering system under scene |
| CN110046244A (en) * | 2019-04-24 | 2019-07-23 | 中国人民解放军国防科技大学 | Answer selection method for question-answering system |
| CN110188175A (en) * | 2019-04-29 | 2019-08-30 | 厦门快商通信息咨询有限公司 | A kind of question and answer based on BiLSTM-CRF model are to abstracting method, system and storage medium |
| CN110246228A (en) * | 2019-05-30 | 2019-09-17 | 中南民族大学 | A kind of chemical experiment evaluation system and its method based on virtual reality technology |
| CN110287294A (en) * | 2018-12-27 | 2019-09-27 | 厦门智融合科技有限公司 | Intellectual property concept answers method and system automatically |
| CN110929265A (en) * | 2019-11-25 | 2020-03-27 | 安徽工业大学 | Multi-angle answer verification method for reading, understanding, asking and answering |
| CN110956142A (en) * | 2019-12-03 | 2020-04-03 | 中国太平洋保险（集团）股份有限公司 | Intelligent interactive training system |
| CN111078875A (en) * | 2019-12-03 | 2020-04-28 | 哈尔滨工程大学 | Method for extracting question-answer pairs from semi-structured document based on machine learning |
| CN112084299A (en) * | 2020-08-05 | 2020-12-15 | 山西大学 | Reading comprehension automatic question-answering method based on BERT semantic representation |
| CN112836034A (en) * | 2021-02-25 | 2021-05-25 | 北京润尼尔网络科技有限公司 | Virtual teaching method and device and electronic equipment |
| CN113420542A (en) * | 2021-06-23 | 2021-09-21 | 平安科技（深圳）有限公司 | Dialog generation method and device, electronic equipment and storage medium |
| CN113722454A (en) * | 2021-07-31 | 2021-11-30 | 东莞理工学院 | Intelligent answering system and method based on admiration course platform |
| WO2022174719A1 (en) * | 2021-02-19 | 2022-08-25 | International Business Machines Corporation | Automatically generating datasets by processing collaboration forums using artificial intelligence techniques |
| CN115331414A (en) * | 2022-10-13 | 2022-11-11 | 江苏濠玥电子科技有限公司 | Temperature early warning method for zirconium oxide ceramic nozzle workpiece |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106569998A (en) * | 2016-10-27 | 2017-04-19 | 浙江大学 | Text named entity recognition method based on Bi-LSTM, CNN and CRF |
| CN107562792A (en) * | 2017-07-31 | 2018-01-09 | 同济大学 | A kind of question and answer matching process based on deep learning |
| CN108108449A (en) * | 2017-12-27 | 2018-06-01 | 哈尔滨福满科技有限责任公司 | A kind of implementation method based on multi-source heterogeneous data question answering system and the system towards medical field |
-
2018
- 2018-06-07 CN CN201810581196.6A patent/CN108804654A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106569998A (en) * | 2016-10-27 | 2017-04-19 | 浙江大学 | Text named entity recognition method based on Bi-LSTM, CNN and CRF |
| CN107562792A (en) * | 2017-07-31 | 2018-01-09 | 同济大学 | A kind of question and answer matching process based on deep learning |
| CN108108449A (en) * | 2017-12-27 | 2018-06-01 | 哈尔滨福满科技有限责任公司 | A kind of implementation method based on multi-source heterogeneous data question answering system and the system towards medical field |
Non-Patent Citations (2)
| Title |
|---|
| 俞可: "面向历史科目的问答技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 张子睿 等: "基于BI-LSTM-CRF模型的中文分词法", 《长春理工大学学报（自然科学版）》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109408627A (en) * | 2018-11-15 | 2019-03-01 | 众安信息技术服务有限公司 | A kind of answering method and system merging convolutional neural networks and Recognition with Recurrent Neural Network |
| CN109739958A (en) * | 2018-11-22 | 2019-05-10 | 普强信息技术（北京）有限公司 | A kind of specification handbook answering method and system |
| CN109671320A (en) * | 2018-12-12 | 2019-04-23 | 广东小天才科技有限公司 | It is a kind of that exercising method and electronic equipment are calculated quickly based on interactive voice |
| CN109671320B (en) * | 2018-12-12 | 2021-06-01 | 广东小天才科技有限公司 | Rapid calculation exercise method based on voice interaction and electronic equipment |
| CN110287294A (en) * | 2018-12-27 | 2019-09-27 | 厦门智融合科技有限公司 | Intellectual property concept answers method and system automatically |
| CN109766427B (en) * | 2019-01-15 | 2021-04-06 | 重庆邮电大学 | Intelligent question-answering method based on collaborative attention for virtual learning environment |
| CN109766427A (en) * | 2019-01-15 | 2019-05-17 | 重庆邮电大学 | A kind of collaborative virtual learning environment intelligent answer method based on stacking Bi-LSTM network and collaboration attention |
| CN109885661A (en) * | 2019-02-27 | 2019-06-14 | 上海优谦智能科技有限公司 | Educate the question answering system under scene |
| CN110046244A (en) * | 2019-04-24 | 2019-07-23 | 中国人民解放军国防科技大学 | Answer selection method for question-answering system |
| CN110046244B (en) * | 2019-04-24 | 2021-06-08 | 中国人民解放军国防科技大学 | Answer selection method for question-answering system |
| CN110188175A (en) * | 2019-04-29 | 2019-08-30 | 厦门快商通信息咨询有限公司 | A kind of question and answer based on BiLSTM-CRF model are to abstracting method, system and storage medium |
| CN110246228A (en) * | 2019-05-30 | 2019-09-17 | 中南民族大学 | A kind of chemical experiment evaluation system and its method based on virtual reality technology |
| CN110246228B (en) * | 2019-05-30 | 2023-03-24 | 中南民族大学 | Chemical experiment evaluation system and method based on virtual reality technology |
| CN110929265B (en) * | 2019-11-25 | 2022-04-26 | 安徽工业大学 | Multi-angle answer verification method for reading, understanding, asking and answering |
| CN110929265A (en) * | 2019-11-25 | 2020-03-27 | 安徽工业大学 | Multi-angle answer verification method for reading, understanding, asking and answering |
| CN110956142A (en) * | 2019-12-03 | 2020-04-03 | 中国太平洋保险（集团）股份有限公司 | Intelligent interactive training system |
| CN111078875A (en) * | 2019-12-03 | 2020-04-28 | 哈尔滨工程大学 | Method for extracting question-answer pairs from semi-structured document based on machine learning |
| CN112084299A (en) * | 2020-08-05 | 2020-12-15 | 山西大学 | Reading comprehension automatic question-answering method based on BERT semantic representation |
| CN112084299B (en) * | 2020-08-05 | 2022-05-31 | 山西大学 | Reading comprehension automatic question-answering method based on BERT semantic representation |
| WO2022174719A1 (en) * | 2021-02-19 | 2022-08-25 | International Business Machines Corporation | Automatically generating datasets by processing collaboration forums using artificial intelligence techniques |
| US11455236B2 (en) | 2021-02-19 | 2022-09-27 | International Business Machines Corporation | Automatically generating datasets by processing collaboration forums using artificial intelligence techniques |
| GB2618953A (en) * | 2021-02-19 | 2023-11-22 | Ibm | Automatically generating datasets by processing collaboration forums using artificial intelligence techniques |
| CN112836034A (en) * | 2021-02-25 | 2021-05-25 | 北京润尼尔网络科技有限公司 | Virtual teaching method and device and electronic equipment |
| CN113420542A (en) * | 2021-06-23 | 2021-09-21 | 平安科技（深圳）有限公司 | Dialog generation method and device, electronic equipment and storage medium |
| CN113420542B (en) * | 2021-06-23 | 2023-04-18 | 平安科技（深圳）有限公司 | Dialogue generation method, device, electronic equipment and storage medium |
| CN113722454A (en) * | 2021-07-31 | 2021-11-30 | 东莞理工学院 | Intelligent answering system and method based on admiration course platform |
| CN115331414A (en) * | 2022-10-13 | 2022-11-11 | 江苏濠玥电子科技有限公司 | Temperature early warning method for zirconium oxide ceramic nozzle workpiece |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108804654A (en) | A kind of collaborative virtual learning environment construction method based on intelligent answer | |
| CN109062939A (en) | A kind of intelligence towards Chinese international education leads method | |
| CN110489755A (en) | Document creation method and device | |
| Freedle | Artificial intelligence and the future of testing | |
| CN106202054A (en) | A kind of name entity recognition method learnt based on the degree of depth towards medical field | |
| CN107544960B (en) | Automatic question-answering method based on variable binding and relation activation | |
| CN111710428B (en) | Biomedical text representation method for modeling global and local context interaction | |
| CN113886567A (en) | Teaching method and system based on knowledge graph | |
| CN116594768A (en) | Large-model-oriented universal tool collaboration and refinement learning system and method | |
| Bagaria et al. | An intelligent system for evaluation of descriptive answers | |
| Gaheen et al. | Automated students arabic essay scoring using trained neural network by e-jaya optimization to support personalized system of instruction | |
| CN112966518B (en) | High-quality answer identification method for large-scale online learning platform | |
| CN113553402B (en) | Automatic question-answering method for reading and understanding examination based on graph neural network | |
| CN116244277A (en) | NLP (non-linear point) identification and knowledge base construction method and system | |
| CN114896975A (en) | Online education intelligent teaching-aid-oriented autonomous evolution method and system | |
| He | Evaluation and stratification for Chinese international education quality with deep learning model | |
| Li | Convolutional neural network-based mining of civic science elements and teaching practice | |
| CN113821610A (en) | Information matching method, device, equipment and storage medium | |
| Shi et al. | Design of personalized recommendation system for swimming teaching based on deep learning | |
| CN117421426B (en) | Knowledge point labeling method and system based on artificial intelligence assistance | |
| Cao et al. | A Study on Game-Based Chinese Vocabulary Learning Tool Powered by Knowledge Graph | |
| Long et al. | Research on Personalized Adaptive Learning System Based on Educational Data Mining | |
| Du | Design and Application of Intelligent Classroom in English Language and Literature Based on Artificial Intelligence Technology | |
| Salem et al. | A concept map based scenario for assessment of short and open answer questions | |
| Xue et al. | An ontology-supported inquiry learning technique |
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 |
Application publication date: 20181113 |
|
| RJ01 | Rejection of invention patent application after publication |