CN112836034A - Virtual teaching method and device and electronic equipment - Google Patents
Virtual teaching method and device and electronic equipment Download PDFInfo
- Publication number
- CN112836034A CN112836034A CN202110222522.6A CN202110222522A CN112836034A CN 112836034 A CN112836034 A CN 112836034A CN 202110222522 A CN202110222522 A CN 202110222522A CN 112836034 A CN112836034 A CN 112836034A
- Authority
- CN
- China
- Prior art keywords
- teaching
- data
- virtual
- deep learning
- virtual teaching
- 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
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
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- 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
- G06F16/338—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/194—Calculation of difference between files
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
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)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Computing Systems (AREA)
- Human Computer Interaction (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The invention provides a virtual teaching method, a virtual teaching device and electronic equipment. The method is applied to a virtual teaching system, wherein the virtual teaching system comprises a plurality of deep learning models, and the deep learning models are obtained by training based on teaching text data; the method comprises the following steps: acquiring virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type; vectorizing the virtual teaching data to obtain vector data; and inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result. In the mode, the virtual teaching system can process virtual teaching data based on a machine learning mode, does not need a teacher to modify, can reduce teaching cost and learning cost, and improves accuracy and efficiency of virtual teaching.
Description
Technical Field
The invention relates to the technical field of machine learning, in particular to a virtual teaching method, a virtual teaching device and electronic equipment.
Background
The virtual simulation experiment integrates a plurality of information technologies, and is a complex system engineering. The early virtual simulation experiment system is oriented to a single subject, the experiment teaching link cannot be fully covered, the experiment process has no intelligent experiment guidance, students have no request for teaching when the teachers are not around, and the experiment results cannot be automatically corrected. The teacher takes the roles of an experiment instructor and a technical service engineer to repeatedly solve various problems of teaching, technology, operation and maintenance and the like brought forward by students. Students also have relative difficulty in learning virtual simulation experiments, and the students are difficult to put efforts on learning simulation experiments due to experimental use problems and technical problems. Experimenters in the virtual space often can produce solitary sense and the sense of nothing help of a certain degree, and the wholesale of experimental result is loaded down with trivial details and low efficiency, and these all restrict the popularization and the popularization of virtual simulation experiment teaching.
Disclosure of Invention
In view of this, the present invention provides a virtual teaching method, a virtual teaching device, and an electronic device, so as to implement a human-machine conversation in virtual teaching, reduce teaching cost and learning cost, and improve accuracy and efficiency of virtual teaching.
In a first aspect, an embodiment of the present invention provides a virtual teaching method, which is applied to a virtual teaching system, where the virtual teaching system includes multiple deep learning models, where the deep learning models are obtained by training based on teaching text data; the method comprises the following steps: acquiring virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type; vectorizing the virtual teaching data to obtain vector data; and inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result.
In a preferred embodiment of the present invention, the virtual education system further includes a distributor; vectorizing the virtual teaching data to obtain vector data, comprising: inputting the virtual teaching data into a distributor, and determining a processing mode of the virtual teaching data; and if the processing mode of the virtual teaching data is a deep learning mode, vectorizing the virtual teaching data to obtain vector data.
In a preferred embodiment of the present invention, the virtual teaching system further includes an expert system model; after the step of determining the processing mode of the virtual teaching data, the method further comprises the following steps: and if the processing mode of the virtual teaching data is the mode of an expert system, inputting the virtual teaching data into an expert system model, and outputting the score corresponding to the virtual teaching data.
In a preferred embodiment of the present invention, the deep learning model includes a first question-answer model based on a search; the teaching type is a question and answer type; inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the method comprises the following steps: and inputting the vector data into the first question-answering model, and outputting a first result.
In a preferred embodiment of the present invention, the deep learning model further includes a second question-answering model based on machine reading understanding; after the step of outputting the first result, the method further comprises: if the degree of association between the first result and the vector data is smaller than a preset first threshold value, inputting the vector data into a second question-answering model, and outputting a second result; or, in response to the modification operation for the first result, inputting the vector data into the second question-answering model, and outputting the second result.
In a preferred embodiment of the present invention, the deep learning model includes a correction model obtained based on teaching text training; the teaching type is a correction type; inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the method comprises the following steps: and inputting the vector data into the correction model to obtain the score corresponding to the virtual teaching data.
In a preferred embodiment of the present invention, the deep learning model includes a guidance model obtained by training using data and scripts based on a virtual teaching system; the teaching type is a guidance type; inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the method comprises the following steps: and inputting the vector data into the guidance model to obtain a guidance flow corresponding to the virtual teaching data.
In a preferred embodiment of the present invention, the virtual teaching system further comprises a custom knowledge base; the method further comprises the following steps: responding to the teaching text adding operation aiming at the user-defined knowledge base, and adding teaching text data in the user-defined knowledge base; and training a plurality of deep learning models based on the custom knowledge base.
In a second aspect, an embodiment of the present invention further provides a virtual teaching device, which is applied to a virtual teaching system, where the virtual teaching system includes a plurality of deep learning models, and the deep learning models are obtained by training based on teaching text data; the device comprises: the data acquisition module is used for acquiring virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type; the vectorization module is used for vectorizing the virtual teaching data to obtain vector data; and the deep learning module is used for inputting the vector data into a deep learning model corresponding to the teaching type and outputting a calculation result.
In a third aspect, an embodiment of the present invention further provides an electronic device, which includes a processor and a memory, where the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the steps of the virtual teaching method described above.
The embodiment of the invention has the following beneficial effects:
according to the virtual teaching method, the virtual teaching device and the electronic equipment, the virtual teaching system can process virtual teaching data based on a machine learning mode without modification by teachers, so that the teaching cost and the learning cost can be reduced, and the accuracy and the efficiency of virtual teaching can be improved.
Additional features and advantages of the disclosure will be set forth in the description which follows, or in part may be learned by the practice of the above-described techniques of the disclosure, or may be learned by practice of the disclosure.
In order to make the aforementioned objects, features and advantages of the present disclosure more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a schematic functional architecture diagram of a virtual teaching system according to an embodiment of the present invention;
fig. 2 is a flowchart of a virtual teaching method according to an embodiment of the present invention;
FIG. 3 is a flowchart of another virtual instruction method according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a model layer of a virtual teaching system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a student question answering according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a virtual teaching apparatus according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
At present, an early virtual simulation experiment system is oriented to a single subject, experiment teaching links cannot be covered completely, an experiment process has no intelligent experiment guidance, students have no request for teaching when teachers are not around, and experiment results cannot be corrected automatically. Experimenters in the virtual space often can produce solitary sense and the sense of nothing help of a certain degree, and the wholesale of experimental result is loaded down with trivial details and low efficiency, and these all restrict the popularization and the popularization of virtual simulation experiment teaching.
Based on this, in order to make artificial intelligence become the medium and the middle bridge that teachers and students accomplished the teaching activity, reduce teaching cost and study cost. The intelligent experimental platform has the advantages that souls are enabled for rigid simulation experiments, the intelligent experimental service platform without subject limitation and experimental type limitation is created, intelligent question answering, intelligent guidance and automatic correction of the experiments are realized, and the intelligent experimental service is opened for users. The embodiment of the invention provides a virtual teaching method, a virtual teaching device and electronic equipment, relates to the technical field of machine learning, natural language processing, voice recognition and education, and particularly relates to an artificial intelligent mixed model correcting and answering system based on virtual simulation.
To facilitate understanding of the embodiment, a detailed description is first given of a virtual teaching method disclosed in the embodiment of the present invention.
The first embodiment is as follows:
the embodiment provides a virtual teaching method, which is applied to a virtual teaching system, wherein the virtual teaching system comprises a plurality of deep learning models, and the deep learning models are obtained by training based on teaching text data.
Referring to a functional architecture schematic diagram of a virtual teaching system shown in fig. 1, the virtual teaching system may be an artificial intelligence hybrid model modifying and answering system based on virtual simulation, and may adopt a hierarchical design, mainly including an application layer, a scheduling layer, a model layer, and a data layer. The application layer is mainly responsible for interacting with the client and receiving and responding to a user request; the scheduling layer is mainly responsible for model concurrent computation so as to respond to multi-user requests; the model layer provides artificial intelligence services such as experiment operation deduction, experiment correction, experiment question and answer, experiment guidance and the like; the data layer is responsible for storing experiment source data, experiment target data and data query service. The design has the advantages that the system is decoupled to the maximum extent, different personnel can be coordinated for parallel development tasks in the later period, the research and development efficiency is improved, and the research and development cost is reduced.
Based on the above description, referring to the flowchart of a virtual teaching method shown in fig. 2, the virtual teaching method includes the following steps:
step S202, acquiring virtual teaching data; the virtual teaching data comprise teaching types, and the teaching types comprise question and answer types, correction types and guidance types.
The virtual teaching data can be data obtained by the operation of a student on the virtual teaching system, and by taking the virtual experiment system as an example, when the student performs an experiment on the virtual experiment system, the virtual experiment system can generate the virtual teaching data according to the operation of the student.
According to different needs of students, different types of teaching can be performed on the students, such as: question answering, correction or instruction. The student or teacher may select one type as the teaching type included in the virtual teaching data.
And step S204, vectorizing the virtual teaching data to obtain vector data.
The virtual teaching data is generally in a text format or a voice format and cannot be directly processed. Therefore, vectorization of the virtual teaching data is required, and the vectorization results in vector data. The vector data can be obtained by directly vectorizing the data in the text format; for the data in the voice format, the data can be converted into the data in the text format, and then vector data can be obtained through vectorization.
And step S206, inputting the vector data into the deep learning model corresponding to the teaching type, and outputting a calculation result.
Different teaching types in this embodiment correspond to different deep learning models, for example: the question-answer type corresponds to a question-answer model, the correction type corresponds to a correction model, and the guidance type corresponds to a guidance model. After the teaching type included in the virtual teaching data is determined, the virtual teaching system can input the vector data into the deep learning model corresponding to the teaching type, and output the calculation result. And, the virtual teaching system can show the calculation result to a teacher or a student for watching.
According to the virtual teaching method provided by the embodiment of the invention, the virtual teaching system can process the virtual teaching data based on a machine learning mode without modification by teachers, so that the teaching cost and the learning cost can be reduced, and the accuracy and the efficiency of virtual teaching can be improved.
Example two:
the embodiment provides another virtual teaching method, which is implemented on the basis of the embodiment; the embodiment focuses on a specific implementation manner of vectorizing the virtual teaching data to obtain vector data. As shown in fig. 3, another flow chart of a virtual teaching method, the virtual teaching method in this embodiment includes the following steps:
step S302, virtual teaching data is obtained; the virtual teaching data comprise teaching types, and the teaching types comprise question and answer types, correction types and guidance types.
When students do experiments in the virtual experiment system, teachers are required to guide when encountering problems, and system prompts are required when the students do not know how to deal with the problems in the next step. After the virtual experiment is finished, a teacher is required to modify the experiment result to give scores and point out detailed scoring points. If the functions are performed by systems instead of human beings, the characteristics of different technologies of artificial intelligence are combined. The expert system can judge according to a definite rule and is suitable for automatically correcting experiments with definite experimental results; deep learning can be used for answering various questions posed by students by training a large amount of corpus knowledge, a knowledge base can be built by self to supplement questions which cannot be answered but cannot be trained, and open questions without fixed answers are also suitable for being judged by adopting a deep learning method. The virtual teaching system provided by the embodiment organically combines the expert system and the deep learning artificial intelligence technology, and solves the problems encountered in the virtual experiment.
Step S304, inputting the virtual teaching data into the distributor, and determining the processing mode of the virtual teaching data.
Referring to fig. 4, a schematic diagram of a model layer of a virtual education system is shown, when a front-end request comes, a distributor determines whether the front-end request is processed by an expert system or a deep learning framework, and then the front-end request is routed to the expert system or the deep learning framework. Specifically, if the virtual teaching data is a test paper of a subjective question, the virtual teaching data can be processed through a deep learning framework, namely the processing mode of the virtual teaching data is a deep learning mode; if the virtual teaching data is an objective test paper, the virtual teaching data can be processed through an expert system framework, namely the processing mode of the virtual teaching data is an expert system mode.
And step S306, if the processing mode of the virtual teaching data is a deep learning mode, vectorizing the virtual teaching data to obtain vector data.
And if the data is routed to the deep learning framework, the requested data is cleaned and vectorized, and then the data enters a model for calculation.
The basic modules of the deep learning framework in the embodiment include a pre-training module: the main data currently processed by the system is unformatted data with language words as contents, because the language word information cannot be directly subjected to numerical calculation, the word information needs to be converted into a high-dimensional space vector of a numerical type before calculation, and the semantic relation of the correlation between words is represented by the similarity between vectors. The vector representation of the characters is obtained through training of a specific corpus, and the text related to disciplines is used, and pre-training is carried out on the basis, so that a pre-training deep learning model containing discipline semantics is obtained.
And step S308, inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result.
For different teaching types, the teaching can be processed through different deep learning models. For example:
the question-answer type (I) is a process that a system gives needed answers according to questions of a user, and at present, a question-answer model based on retrieval and a question-answer model based on machine reading are mainly applied to the system, and the two models are combined to serve. The retrieval model is based on a question answer to corpus, the system converts the questions of the user into numerical vectors of the questions according to a pre-training model, then calculates the semantic relevance between the received questions and the questions collected in the corpus, and the system feeds the answers corresponding to the questions with the highest relevance back to the user according to the semantic relevance of the questions.
(1) The deep learning model comprises a first question-answer model based on retrieval; the teaching type is a question and answer type; can be processed in the following way: and inputting the vector data into the first question-answering model, and outputting a first result.
And (3) realizing a question-answer model based on retrieval: and (4) collecting question-answer pairs to form a question-answer-pair corpus, matching the questions in the corpus according to the numerical vector representation of the questions in the pre-training model, and returning answers to the matched questions.
If the user's question is generally associated with a low degree of question relevance in the corpus or the user is dissatisfied with the retrieved answer, the system will use a question-answer model based on reading understanding, and the reading understanding-based model will extract relevant content from the subject text data as the answer according to the question.
(2) The deep learning model further comprises a second question-answering model based on machine reading understanding; if the degree of association between the first result and the vector data is smaller than a preset first threshold value, inputting the vector data into a second question-answering model, and outputting a second result; or, in response to the modification operation for the first result, inputting the vector data into the second question-answering model, and outputting the second result.
The realization of the question-answering model based on machine reading understanding: the positions of the answers appearing in the subject text are labeled, and then a reading understanding model is trained according to the questions and the labeled text. When the system receives a question, a paragraph set possibly containing answers is determined according to the question, then the positions of the answers in the paragraphs are determined in the texts according to a machine reading model, and the answers are extracted according to the corresponding positions.
According to the user feedback and the retrieval result, the retrieval model or the reading-based question-answer model is comprehensively judged and used, so that a method for combining the retrieval-based question-answer model and the machine-reading-based question-answer model is realized, the question-answer function is more complete, the model is gradually updated according to data gradually accumulated by the system after the system is deployed on line, and the accuracy and the user satisfaction are improved.
(II) correcting types, wherein the deep learning model comprises a correcting model obtained based on teaching text training; the teaching type is a correction type and can be processed by the following steps: and inputting the vector data into the correction model to obtain the score corresponding to the virtual teaching data.
The method includes the steps that texts submitted by students, such as experimental reports, open question answers, standard texts given by teachers or product managers, or fixed paragraphs extracted from subject materials according to questions, the texts without the fixed answers are scored by using text editing methods, similarity calculation methods and the like, and parts which do not meet experimental requirements or question requirements are pointed out.
(III) guiding types, wherein the deep learning model comprises a guiding model obtained by training based on the use data and the use script of the virtual teaching system; the teaching type is a guidance type; can be processed by the following steps: inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the method comprises the following steps: and inputting the vector data into the guidance model to obtain a guidance flow corresponding to the virtual teaching data.
For the experiment with higher openness, namely the experiment without a specific flow, similarity calculation is carried out according to the operation data script of the user and the collected operation script set to obtain an operation script sequence with higher similarity, and then the operation scheme in the collected script is used as an operation guidance scheme given to the user by the system and is fed back to the user. And for the experiment with a specific flow, a rule engine is used for guiding, and a specific guide flow is given.
In this embodiment, the deep learning model may be trained through a custom knowledge base, wherein the custom knowledge base may be added with teaching text data by a teacher, for example: responding to the teaching text adding operation aiming at the user-defined knowledge base, and adding teaching text data in the user-defined knowledge base; and training a plurality of deep learning models based on the custom knowledge base. Teachers can customize the knowledge base to supplement professional knowledge, and answer accuracy is improved.
And S310, if the processing mode of the virtual teaching data is the mode of an expert system, inputting the virtual teaching data into an expert system model, and outputting the score corresponding to the virtual teaching data.
If routed to the expert system, the expert system performs calculations according to preset rules using a rule engine. And after the calculation is finished, returning the result to the front-end caller.
For example, the virtual teaching system provided in this embodiment can be used through the following steps (1) to (5):
and (1) constructing a pre-training model.
Collecting subject text data related to middle school and university experiments, and mainly analyzing related course knowledge; the course data is used as the basic data of pre-training, and on the basis, a language model training method in a natural language processing method is utilized to train a pre-training language model specially aiming at subject data. In the training process, different neural network models are tried to be used for testing and comparing the training effect, and the model with the best effect is used as a pre-training language model to be used and is mainly used for converting text characters into numerical vectors which can be calculated by the neural network.
And (2) building a question-answer model.
1) Question-answer model based on retrieval: firstly, collecting question answer pairs related to subject experiments to form a knowledge base with the question answer pairs as main contents; secondly, converting the problem into corresponding word vector representation by utilizing a pre-trained language model; then, a method for calculating the similarity between the user question and the system included question is utilized to obtain a corresponding answer; and finally, according to the requirements of teachers, students and product managers, data are continuously added, the model is iteratively optimized, the accuracy of question answering is improved, and the range of the answers can be expanded.
2) Question-answer model based on machine reading: firstly, collecting questions and subject text corpora related to the questions, namely finding related texts containing the questions; secondly, marking the range of the answers of the questions in the related texts, wherein the marking process is completed by mutually combining manual marking and machine marking, machine marking is carried out on the questions with collected answers in the marking process, and manual marking is carried out on the questions without answers; then, training the related reading understanding model by using the marked data and a training process of a common machine reading understanding model, and deploying after completing testing and evaluation; and finally, after the model is on line, performing iterative training by using newly added data, optimizing the model and improving the accuracy.
And (3) constructing a correction model.
Firstly, collecting open questions and answer texts with standard comparison; then, establishing a grading model and a correcting model aiming at the open problems by using the collected problems and text data and using methods of text editing, similarity calculation, text sequencing and the like, wherein the grading model is mainly used for carrying out score evaluation on contents, and the correcting model is used for correcting texts submitted by students; and finally, changing the model according to the feedback of the students and teachers on the correction and grading results, and improving the user experience.
And (4) guiding the construction of the model.
Firstly, a product manager and a teacher make a relatively complete experimental process in a front-end virtual experimental system, then an experiment is submitted, and an experimental process script is collected by the system; secondly, analyzing the collected script, extracting experiment operation information contained in the script and forming an experiment operation sequence library; then, establishing a guidance model aiming at the experiment with higher openness according to the operation sequence library; and finally, performing iterative optimization on the guidance model.
And (5) building a rule base.
A teacher completely completes an experiment process in a front-end virtual experiment system, then submits the experiment, the system collects an experiment result script and an experiment process script formed by the teacher doing the experiment, and rules are automatically extracted to form a rule base by analyzing the experiment result script and the experiment process script.
In addition, referring to a schematic diagram of student question answering shown in fig. 5, when a student clicks a question-asking button in an experimental process, a dialog box is popped out in a virtual experimental scene, a user can use character input or click a microphone icon to carry out voice question answering, and a system can provide text answers and voice answers of questions. In the experimental process, the student clicks a button for requesting guidance, and then the system gives corresponding next step guidance or corrects the wrong step of the student according to the current experimental process of the student. When the student submits the experiment after finishing the experiment, the system can carry out the correction by oneself and give the student's experimental score.
The virtual teaching system that this embodiment provided has adopted the mode that expert system and deep learning combined together, has trained multiple model combination and has used under the deep learning frame, has realized functions such as the correction answer on the virtual experiment, teaching guidance, and the virtual teaching system that this embodiment provided has following characteristics:
(1) two artificial intelligence implementation modes of an expert system and deep learning are integrated.
(2) Human-machine conversation in virtual experiments is achieved.
(3) Teachers can customize the knowledge base to supplement professional knowledge, and answer accuracy is improved.
(4) The combination of different models is deeply learned, the performance of the system is improved, and a scheme is provided for better improving the performance of the deep learning model.
(5) And the knowledge integration is realized by integrating the multidisciplinary knowledge and the multi-field knowledge, and the fragmented knowledge is integrated to form a complete knowledge system, so that richer and systematized related field knowledge is provided for the education products.
It should be noted that the above method embodiments are all described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other.
Example three:
corresponding to the method embodiment, the embodiment of the invention provides a virtual teaching device, which is applied to a virtual teaching system, wherein the virtual teaching system comprises a plurality of deep learning models, and the deep learning models are obtained based on teaching text data training; fig. 6 is a schematic structural diagram of a virtual teaching device, which includes:
a data obtaining module 61, configured to obtain virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type;
a vectorization module 62, configured to vectorize the virtual teaching data to obtain vector data;
and the deep learning module 63 is configured to input the vector data into a deep learning model corresponding to the teaching type, and output a calculation result.
According to the virtual teaching device provided by the embodiment of the invention, the virtual teaching system can process the virtual teaching data based on a machine learning mode without modification by teachers, so that the teaching cost and the learning cost can be reduced, and the accuracy and the efficiency of virtual teaching can be improved.
The virtual teaching system further comprises a distributor; the vectorization module is used for inputting the virtual teaching data into the distributor and determining the processing mode of the virtual teaching data; and if the processing mode of the virtual teaching data is a deep learning mode, vectorizing the virtual teaching data to obtain vector data.
The virtual teaching system also comprises an expert system model; the device further comprises an expert system and a module, wherein the module is used for inputting the virtual teaching data into the expert system model and outputting the scores corresponding to the virtual teaching data if the processing mode of the virtual teaching data is the expert system mode.
The deep learning model comprises a first question-answer model based on retrieval; the teaching type is a question and answer type; the deep learning module is used for inputting the vector data into the first question-answering model and outputting a first result.
The deep learning model further comprises a second question-answering model based on machine reading understanding; the deep learning module is further configured to input the vector data into a second question-answering model and output a second result if the association degree between the first result and the vector data is smaller than a preset first threshold; or, in response to the modification operation for the first result, inputting the vector data into the second question-answering model, and outputting the second result.
The deep learning model comprises a correction model obtained based on teaching text training; the teaching type is a correction type; the deep learning module is used for inputting the vector data into the correction model to obtain the score corresponding to the virtual teaching data.
The deep learning model comprises a guiding model obtained by training using data and using scripts based on a virtual teaching system; the teaching type is a guidance type; the deep learning module is used for inputting the vector data into the guidance model to obtain a guidance flow corresponding to the virtual teaching data.
The virtual teaching system also comprises a user-defined knowledge base; the above-mentioned device still includes: the knowledge base adding module is used for responding to the teaching text adding operation aiming at the user-defined knowledge base and adding teaching text data in the user-defined knowledge base; and training a plurality of deep learning models based on the custom knowledge base.
Example four:
the embodiment of the invention also provides electronic equipment, which is used for operating the virtual teaching method; referring to fig. 7, a schematic structural diagram of an electronic device is shown, the electronic device includes a memory 100 and a processor 101, where the memory 100 is used for storing one or more computer instructions, and the one or more computer instructions are executed by the processor 101 to implement the virtual teaching method.
Further, the electronic device shown in fig. 7 further includes a bus 102 and a communication interface 103, and the processor 101, the communication interface 103, and the memory 100 are connected through the bus 102.
The Memory 100 may include a high-speed Random Access Memory (RAM) and may further include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. The communication connection between the network element of the system and at least one other network element is realized through at least one communication interface 103 (which may be wired or wireless), and the internet, a wide area network, a local network, a metropolitan area network, and the like can be used. The bus 102 may be an ISA bus, PCI bus, EISA bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 7, but this does not indicate only one bus or one type of bus.
The processor 101 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware or instructions in the form of software in the processor 101. The Processor 101 may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the device can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in the memory 100, and the processor 101 reads the information in the memory 100, and completes the steps of the method of the foregoing embodiment in combination with the hardware thereof.
The embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are called and executed by a processor, the computer-executable instructions cause the processor to implement the virtual teaching method.
The virtual teaching method, the virtual teaching device, and the computer program product of the electronic device provided in the embodiments of the present invention include a computer-readable storage medium storing a program code, where instructions included in the program code may be used to execute the method in the foregoing method embodiments, and specific implementation may refer to the method embodiments, and will not be described herein again.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and/or the apparatus described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In addition, in the description of the embodiments of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A virtual teaching method is characterized by being applied to a virtual teaching system, wherein the virtual teaching system comprises a plurality of deep learning models, and the deep learning models are obtained by training based on teaching text data; the method comprises the following steps:
acquiring virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type;
vectorizing the virtual teaching data to obtain vector data;
and inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result.
2. The method of claim 1, wherein the virtual tutorial system further comprises a dispatcher; vectorizing the virtual teaching data to obtain vector data, wherein the vectorizing comprises the following steps:
inputting the virtual teaching data into the distributor, and determining a processing mode of the virtual teaching data;
and if the processing mode of the virtual teaching data is a deep learning mode, vectorizing the virtual teaching data to obtain vector data.
3. The method of claim 2, wherein the virtual tutorial system further comprises an expert system model; after the step of determining the processing mode of the virtual teaching data, the method further comprises:
and if the processing mode of the virtual teaching data is an expert system mode, inputting the virtual teaching data into the expert system model and outputting the score corresponding to the virtual teaching data.
4. The method of claim 1, wherein the deep learning model comprises a first question-answer model based on a search; the teaching type is the question and answer type; inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the step comprises the following steps of:
and inputting the vector data into the first question-answering model, and outputting a first result.
5. The method of claim 4, wherein the deep learning model further comprises a second question-and-answer model based on machine-reading understanding; after the step of outputting the first result, the method further comprises:
if the association degree of the first result and the vector data is smaller than a preset first threshold value, inputting the vector data into the second question-answering model, and outputting a second result;
or, in response to the modification operation for the first result, inputting the vector data into the second question-answering model, and outputting a second result.
6. The method of claim 1, wherein the deep learning model comprises a correction model trained based on instructional text; the teaching type is the correcting type;
inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the step comprises the following steps of:
and inputting the vector data into the correction model to obtain a score corresponding to the virtual teaching data.
7. The method of claim 1, wherein the deep learning model comprises a guidance model trained based on usage data and usage scripts of the virtual tutorial system; the teaching type is the guidance type;
inputting the vector data into a deep learning model corresponding to the teaching type, and outputting a calculation result, wherein the step comprises the following steps of:
and inputting the vector data into the guidance model to obtain a guidance flow corresponding to the virtual teaching data.
8. The method of claim 1, wherein the virtual tutoring system further comprises a custom knowledge base; the method further comprises the following steps:
responding to a teaching text adding operation aiming at the user-defined knowledge base, and adding teaching text data in the user-defined knowledge base;
training a plurality of the deep learning models based on the custom knowledge base.
9. A virtual teaching device is applied to a virtual teaching system, wherein the virtual teaching system comprises a plurality of deep learning models, and the deep learning models are obtained by training based on teaching text data; the device comprises:
the data acquisition module is used for acquiring virtual teaching data; the virtual teaching data comprise teaching types, wherein the teaching types comprise a question and answer type, a correction type and a guidance type;
the vectorization module is used for vectorizing the virtual teaching data to obtain vector data;
and the deep learning module is used for inputting the vector data into a deep learning model corresponding to the teaching type and outputting a calculation result.
10. An electronic device comprising a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to perform the steps of the virtual teaching method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110222522.6A CN112836034A (en) | 2021-02-25 | 2021-02-25 | Virtual teaching method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110222522.6A CN112836034A (en) | 2021-02-25 | 2021-02-25 | Virtual teaching method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112836034A true CN112836034A (en) | 2021-05-25 |
Family
ID=75934002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110222522.6A Pending CN112836034A (en) | 2021-02-25 | 2021-02-25 | Virtual teaching method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112836034A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108804654A (en) * | 2018-06-07 | 2018-11-13 | 重庆邮电大学 | A kind of collaborative virtual learning environment construction method based on intelligent answer |
| CN109271505A (en) * | 2018-11-12 | 2019-01-25 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on problem answers pair |
| CN109766423A (en) * | 2018-12-29 | 2019-05-17 | 上海智臻智能网络科技股份有限公司 | Answering method and device neural network based, storage medium, terminal |
| 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 |
| CN110020009A (en) * | 2017-09-29 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Online answering method, apparatus and system |
| CN111552773A (en) * | 2020-04-24 | 2020-08-18 | 中国科学院空天信息创新研究院 | Method and system for searching key sentence of question or not in reading and understanding task |
-
2021
- 2021-02-25 CN CN202110222522.6A patent/CN112836034A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020009A (en) * | 2017-09-29 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Online answering method, apparatus and system |
| CN108804654A (en) * | 2018-06-07 | 2018-11-13 | 重庆邮电大学 | A kind of collaborative virtual learning environment construction method based on intelligent answer |
| CN109271505A (en) * | 2018-11-12 | 2019-01-25 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on problem answers pair |
| CN109766423A (en) * | 2018-12-29 | 2019-05-17 | 上海智臻智能网络科技股份有限公司 | Answering method and device neural network based, storage medium, terminal |
| 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 |
| CN111552773A (en) * | 2020-04-24 | 2020-08-18 | 中国科学院空天信息创新研究院 | Method and system for searching key sentence of question or not in reading and understanding task |
Non-Patent Citations (1)
| Title |
|---|
| 乔非等: "《虚拟教学实验设计和开发技术与应用》", 30 December 2013 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110362671B (en) | Topic recommendation method, device and storage medium | |
| CN101673356B (en) | Method and system for generating dialogue managers with diversified dialogue acts | |
| CN110797010A (en) | Question-answer scoring method, device, equipment and storage medium based on artificial intelligence | |
| CN111310463B (en) | Test question difficulty estimation method and device, electronic equipment and storage medium | |
| US11989524B2 (en) | Knowledge-grounded dialogue system and method for language learning | |
| CN112685550B (en) | Intelligent question-answering method, intelligent question-answering device, intelligent question-answering server and computer readable storage medium | |
| CN109949637A (en) | A kind of objective item purpose answers method and apparatus automatically | |
| CN106897384A (en) | One kind will bring out the theme automatic evaluation method and device | |
| CN110991195A (en) | Machine translation model training method, device and storage medium | |
| CN117077792B (en) | Knowledge graph-based method and device for generating prompt data | |
| CN114218379A (en) | Intelligent question-answering system-oriented method for attributing questions which cannot be answered | |
| CN112506945A (en) | Self-adaptive learning guiding method and system based on knowledge graph | |
| CN115630860A (en) | Teaching quality evaluation method, device, equipment and storage medium | |
| CN110765241A (en) | Super-outline detection method and device for recommendation questions, electronic equipment and storage medium | |
| CN117150151B (en) | Wrong question analysis and test question recommendation system and method based on large language model | |
| CN111625631B (en) | Method for generating option of choice question | |
| CN117235347A (en) | Teenager algorithm code aided learning system and method based on large language model | |
| CN116362331A (en) | Knowledge point filling method based on man-machine cooperation construction knowledge graph | |
| CN112836034A (en) | Virtual teaching method and device and electronic equipment | |
| CN116401341A (en) | Interactive answering system oriented to understanding | |
| CN113821610A (en) | Information matching method, device, equipment and storage medium | |
| CN114358579A (en) | Evaluation method, evaluation device, electronic device, and computer-readable storage medium | |
| CN113569112A (en) | Tutoring strategy providing method, system, device and medium based on question | |
| KR20220118579A (en) | System for providing tutoring service using artificial intelligence and method thereof | |
| CN112256743A (en) | Adaptive question setting method, equipment and storage medium |
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: 20210525 |
|
| RJ01 | Rejection of invention patent application after publication |