CN115510857A - Game education knowledge graph construction method and device - Google Patents
Game education knowledge graph construction method and device Download PDFInfo
- Publication number
- CN115510857A CN115510857A CN202211467546.9A CN202211467546A CN115510857A CN 115510857 A CN115510857 A CN 115510857A CN 202211467546 A CN202211467546 A CN 202211467546A CN 115510857 A CN115510857 A CN 115510857A
- Authority
- CN
- China
- Prior art keywords
- game
- game element
- neural network
- training
- elements
- 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
- 238000010276 construction Methods 0.000 title claims abstract description 24
- 238000013528 artificial neural network Methods 0.000 claims abstract description 39
- 238000003062 neural network model Methods 0.000 claims description 70
- 238000000034 method Methods 0.000 claims description 22
- 230000011218 segmentation Effects 0.000 claims description 10
- 238000005457 optimization Methods 0.000 claims description 8
- 230000018109 developmental process Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The application provides a game education knowledge graph construction method and device, at least one vocabulary element is obtained by segmenting text data of a game chat room, corresponding eigenvectors are constructed for all the vocabulary elements, game element recognition is carried out on all the vocabulary elements, game element contact judgment is carried out on all the game elements according to the eigenvectors corresponding to all the game elements and a preset game element contact neural network, game element contact conditions among all the game elements are obtained, and a game education knowledge graph is constructed according to all the game elements, marks corresponding to all the game elements and game element contact conditions among all the game elements.
Description
Technical Field
The invention relates to the field of game education, in particular to a game education knowledge graph construction method and device.
Background
With the development of times society, people's understanding of games has changed day by day, and most people do not talk about the change of game colors. Meanwhile, games play an increasingly important role in the development of the economic society. Under such circumstances, many schools have developed game-related specialties. On the one hand, the game professional is a new-born military in the education industry, and the related knowledge of the game professional is lacked relative to other mature professionals in the prior art, which is not beneficial to the development of the game and the game professional. On the other hand, the games are various, the game-related materials are also visible anywhere on the network, and the game education knowledge graph is feasible to be constructed by using the game-related materials.
Disclosure of Invention
A game education knowledge graph construction method comprises the following steps:
segmenting words of the text data of the game chat room to obtain at least one word element;
constructing corresponding eigenvectors for all the vocabulary elements;
according to the eigenvectors corresponding to all the vocabulary elements and a preset game element judgment model, carrying out game element identification on all the vocabulary elements, and marking all the identified game elements;
performing game element contact judgment on all the game elements according to the eigenvectors corresponding to all the game elements and a preset game element contact neural network to obtain game element contact conditions among all the game elements;
and constructing a game education knowledge graph according to all the game elements, the marks corresponding to all the game elements and the game element contact conditions among all the game elements.
Preferably, the specific process of presetting the game element determination model includes:
segmenting a text data game element training corpus of a game chat room to obtain at least one game element training vocabulary element;
constructing corresponding game element training eigenvectors for all the game element training vocabulary elements;
taking all the game element training eigenvectors as input data of a first BERT-based neural network model for identifying game elements, and carrying out layer-by-layer unsupervised pre-training on the first BERT-based neural network model, wherein the first BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
adding a neural network layer as an output layer in the pre-trained first BERT-based neural network model to obtain a trained first BERT-based neural network model;
according to standard game element marks corresponding to all the game element training eigenvectors, optimizing game element identification parameters of each layer of the trained first BERT-based neural network model based on Backpropagation;
and determining the trained first BERT-based neural network model optimized based on Backpropagation as a game element judgment model.
Preferably, the specific process of presetting the game element-to-neural network includes:
segmenting words of a game chatting room text data game element association training corpus to obtain at least one game element association training vocabulary element;
constructing corresponding game element association training eigenvectors for all the game element association training vocabulary elements;
taking all game element association training eigenvectors as input data of a second BERT-based neural network model for identifying game element association, and performing layer-by-layer unsupervised pre-training on the second BERT-based neural network model, wherein the second BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
adding a neural network layer as an output layer in the pre-trained second BERT-based neural network model to obtain a second BERT-based neural network model;
according to standard game element connection marks among all the game element connection training eigenvectors, optimizing game element connection judgment parameters of each layer of the second BERT-based neural network model based on Backpropagation;
and determining the second BERT-based neural network model optimized based on Backpropagation as a game element contact neural network.
Preferably, the constructing a game education knowledge graph according to all the game elements, the marks corresponding to all the game elements and the game element association conditions among all the game elements includes:
respectively determining all the game elements and the corresponding marks thereof as chart points, and respectively determining the game element connection conditions among all the game elements as sides;
and constructing a game education knowledge graph according to all the graph points and all the edges.
A game education knowledge map construction apparatus comprising: the system comprises a word segmentation module, an eigenvector construction module, a game element identification module, a game element association judgment module and a game education knowledge map construction module; the game element identification module comprises a game element judgment model generation unit, and the game element association judgment module comprises a game element association neural network generation unit;
the word segmentation model is used for segmenting the text data of the game chat room to obtain at least one word element;
the eigenvector construction module is used for constructing corresponding eigenvectors for all the vocabulary elements;
the game element identification module is used for identifying game elements of all the vocabulary elements according to the eigenvectors corresponding to all the vocabulary elements and a preset game element judgment model, and marking all the identified game elements;
the game element contact judgment module is used for carrying out game element contact judgment on all the game elements according to the eigenvectors corresponding to all the game elements and a preset game element contact neural network to obtain game element contact conditions among all the game elements;
the game education knowledge map building module is used for building a game education knowledge map according to all the game elements, the marks corresponding to all the game elements and the game element contact conditions among all the game elements;
the game element judgment model generation unit is used for presetting a game element judgment model;
and the game element contact neural network generating unit is used for presetting a game element contact neural network.
Preferably, the game play element determination model generation means includes: the system comprises a first sub-word module, a first eigenvector building sub-module, a first pre-training sub-module, a first adding sub-module, a first backspace-based optimization sub-module and a game element judgment model determining sub-module;
the first word segmentation submodule is used for segmenting words of the game element training corpus of the text data of the game chat room to obtain at least one game element training vocabulary element;
the first eigenvector construction submodule is used for constructing corresponding game element training eigenvectors for all the game element training vocabulary elements;
the first pre-training sub-module is used for taking all the game element training eigenvectors as input data of a first BERT-based neural network model for identifying game elements and performing layer-by-layer unsupervised pre-training on the first BERT-based neural network model, wherein the first BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
the first adding submodule is used for adding a neural network layer in the pre-trained first BERT-based neural network model to serve as an output layer to obtain a trained first BERT-based neural network model;
the first backward propagation-based optimization submodule is used for optimizing game element identification parameters of each layer of the trained first BERT-based neural network model based on backward propagation according to standard game element marks corresponding to all the game element training eigenvectors;
and the game element judgment model determining submodule is used for determining the trained first BERT-based neural network model optimized based on the Backpropagation as a game element judgment model.
Preferably, the game element contact neural network generating unit includes: the system comprises a second word sub-module, a second eigenvector building sub-module, a second pre-training sub-module, a second adding sub-module, a second backspace-based optimization sub-module and a game element association neural network determining sub-module;
the second word segmentation submodule is used for segmenting words of the game chatting room text data game element association training corpus to obtain at least one game element association training vocabulary element;
the second eigenvector construction submodule is used for constructing corresponding game element association training eigenvectors for all the game element association training vocabulary elements;
the second pre-training sub-module is used for taking all game element association training eigenvectors as input data of a second BERT-based neural network model for identifying game element association and performing layer-by-layer unsupervised pre-training on the second BERT-based neural network model, and the second BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
the second adding submodule is used for adding a neural network layer in the pre-trained second BERT-based neural network model to serve as an output layer to obtain a second BERT-based neural network model;
the second backward propagation-based optimization submodule is used for optimizing game element connection judgment parameters of each layer of the second BERT-based neural network model based on backward propagation according to standard game element connection marks among all game element connection training eigenvectors;
and the game element contact neural network determining submodule is used for determining the second BERT-based neural network model optimized based on Backpropagation as a game element contact neural network.
Preferably, the game education knowledge graph building module includes: a determining unit and a constructing unit;
the determining unit is used for determining all the game elements and the marks corresponding to the game elements as chart points respectively, and determining the game element relation conditions among all the game elements as sides respectively;
and the construction unit is used for constructing the game education knowledge graph according to all the graph points and all the edges.
Drawings
FIG. 1 is a flow chart of a method for constructing a knowledge graph of game education according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention discloses a game education knowledge graph construction method, a flow chart of the method is shown in figure 1, and the method comprises the following steps:
step101, performing word segmentation on the text data of the game chat room to obtain at least one word element;
step102, constructing corresponding eigenvectors for all vocabulary elements;
step103, identifying all the vocabulary elements according to the eigenvectors corresponding to all the vocabulary elements and a preset game element judgment model, and marking all the identified game elements;
in the process of executing Step103, all the recognized game elements are marked.
Step104, performing game element association judgment on all game elements according to the eigenvectors corresponding to all the game elements and a preset game element association neural network to obtain game element association conditions among all the game elements;
step105, constructing a game education knowledge graph according to all game elements, marks corresponding to all game elements and game element connection conditions among all game elements;
optionally, the Step105 includes the following steps:
s201, respectively determining all game elements and corresponding marks thereof as chart points, and respectively determining game element connection conditions among all game elements as sides;
and S202, constructing the game education knowledge graph according to all the graph points and all the edges.
According to the game education knowledge graph construction method provided by the embodiment of the invention, the game education knowledge graph corresponding to the text data of the game chat room is constructed through the preset game element judgment model and the game element contact neural network. Therefore, the learned model can be directly applied to game chat room text data game elements and game element contact judgment in the field of games, and the problem that the existing shallow learning method or pattern matching method is inconvenient for expanding the content of a game system and cannot provide visual reference for game planners is solved.
Example two
With reference to the method for constructing a game education knowledge graph disclosed in the first embodiment of the present invention, as shown in Step103 in fig. 1, a specific process of a game element determination model is preset, including the following steps:
step301, performing word segmentation on the text data game element training corpus of the game chat room to obtain at least one game element training vocabulary element;
step302, training vocabulary elements of all game elements, and constructing corresponding training eigenvectors of the game elements;
step303, taking all game element training eigenvectors as input data of a first BERT-based neural network model for identifying game elements, and training the first BERT-based neural network model, wherein the first BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
step304, adding a neural network layer as an output layer in the pre-trained first BERT-based neural network model to obtain the trained first BERT-based neural network model;
in the process of executing Step304, the specific training process is described as the following processes of steps S305 and S306.
Step305, according to standard game element marks corresponding to all game element training eigenvectors, optimizing and training game element identification parameters of each layer of the first BERT-based neural network model based on Backpropagation;
step306, determining the trained first BERT-based neural network model optimized based on Backpropagation as a game element judgment model.
According to the game education knowledge graph construction method provided by the embodiment of the invention, the game education knowledge graph corresponding to the text data of the game chat room is constructed through the preset game element judgment model and the game element contact neural network.
With reference to the method for constructing a game education knowledge graph disclosed in the embodiment of the present invention, as shown in Step104 in fig. 1, a specific implementation process of a game element-to-neural network is preset, which includes the following steps:
step401, performing word segmentation on the game chatting room text data game element association training corpus to obtain at least one game element association training vocabulary element;
step402, establishing corresponding game element association training eigenvectors for all game element association training vocabulary elements;
step403, taking all game element association training eigenvectors as input data of a second BERT-based neural network model for identifying game element association, and training the second BERT-based neural network model, wherein the second BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer;
step404, adding a neural network layer as an output layer in the pre-trained second BERT-based neural network model to obtain a second BERT-based neural network model;
in the process of executing Step304, the specific training process is described as the following steps S405 and S406.
Step405, optimizing game element association judgment parameters of each layer of a second BERT-based neural network model based on Backpropagation according to standard game element association marks among all game element association training eigenvectors;
and Step406, determining the second BERT-based neural network model optimized based on Backpropagation as a game element contact neural network.
According to the game education knowledge graph construction method provided by the embodiment of the invention, the game education knowledge graph corresponding to the text data based on the game chat room is constructed through the preset game element judgment model and the game element contact neural network.
Claims (7)
1. A game education knowledge graph construction method is characterized by comprising the following steps: segmenting words of the text data of the game chat room to obtain at least one word element; constructing corresponding eigenvectors for all the vocabulary elements; according to the eigenvectors corresponding to all the vocabulary elements and a preset game element judgment model, carrying out game element identification on all the vocabulary elements, and marking all the identified game elements; performing game element contact judgment on all the game elements according to the eigenvectors corresponding to all the game elements and a preset game element contact neural network to obtain game element contact conditions among all the game elements; and constructing a game education knowledge graph according to all the game elements, the marks corresponding to all the game elements and the game element contact conditions among all the game elements.
2. The method of claim 1, wherein the specific process of the predetermined game element determination model comprises: segmenting words of a game chatting room text data game element training corpus to obtain at least one game element training vocabulary element; constructing corresponding game element training eigenvectors for all the game element training vocabulary elements; taking all the game element training eigenvectors as input data of a first BERT-based neural network model for identifying game elements, and performing layer-by-layer unsupervised pre-training on the first BERT-based neural network model, wherein the first BERT-based neural network model is formed by overlapping at least one multi-point attention mechanism layer; adding a neural network layer as an output layer in the pre-trained first BERT-based neural network model to obtain a trained first BERT-based neural network model; according to standard game element marks corresponding to all the game element training eigenvectors, optimizing game element identification parameters of each layer of the trained first BERT-based neural network model based on Backpropagation; and determining the trained first BERT-based neural network model optimized based on Backpropagation as a game element judgment model.
3. The method of claim 1, wherein the presetting of game element-to-neural network specific procedures comprises: segmenting words of a game chatting room text data game element association training corpus to obtain at least one game element association training vocabulary element; constructing corresponding game element association training eigenvectors for all the game element association training vocabulary elements; taking all game element association training eigenvectors as input data of a second BERT-based neural network model for identifying game element association, and performing layer-by-layer unsupervised pre-training on the second BERT-based neural network model, wherein the second BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer; adding a neural network layer as an output layer in the pre-trained second BERT-based neural network model to obtain a second BERT-based neural network model; optimizing game element contact judgment parameters of each layer of the second BERT-based neural network model based on Backpropagation according to standard game element contact marks among all the game element contact training eigenvectors; and determining the second BERT-based neural network model optimized based on Backpropagation as a game element contact neural network.
4. The method of claim 1, wherein said constructing a game education knowledge graph based on all of said game elements, said labels corresponding to all of said game elements, and said game element associations between all of said game elements comprises: respectively determining all the game elements and the corresponding marks thereof as chart points, and respectively determining the game element connection conditions among all the game elements as sides; and constructing a game education knowledge graph according to all the graph points and all the edges.
5. A game education knowledge map construction apparatus characterized by comprising: the system comprises a word segmentation module, an eigenvector construction module, a game element identification module, a game element association judgment module and a game education knowledge map construction module; the game element identification module comprises a game element judgment model generation unit, and the game element association judgment module comprises a game element association neural network generation unit; the word segmentation model is used for segmenting the text data of the game chat room to obtain at least one word element; the eigenvector construction module is used for constructing corresponding eigenvectors for all the vocabulary elements; the game element recognition module is used for recognizing game elements of all the vocabulary elements according to the eigenvectors corresponding to all the vocabulary elements and a preset game element judgment model, and marking all the recognized game elements; the game element association judgment module is used for performing game element association judgment on all the game elements according to the eigenvectors corresponding to all the game elements and a preset game element association neural network to obtain game element association conditions among all the game elements; the game education knowledge map building module is used for building a game education knowledge map according to all the game elements, the marks corresponding to all the game elements and the game element contact conditions among all the game elements; the game element judgment model generation unit is used for presetting a game element judgment model; and the game element contact neural network generating unit is used for presetting a game element contact neural network.
6. The apparatus according to claim 5, wherein the game play element determination model generation means includes: the system comprises a first sub-word module, a first eigenvector building sub-module, a first pre-training sub-module, a first adding sub-module, a first backspace-based optimization sub-module and a game element judgment model determining sub-module; the first word segmentation submodule is used for segmenting words of the game element training corpus of the text data of the game chat room to obtain at least one game element training vocabulary element; the first eigenvector construction submodule is used for constructing corresponding game element training eigenvectors for all the game element training vocabulary elements; the first pre-training sub-module is used for taking all the game element training eigenvectors as input data of a first BERT-based neural network model for identifying game elements and performing layer-by-layer unsupervised pre-training on the first BERT-based neural network model, wherein the first BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer; the first adding submodule is used for adding a neural network layer in the pre-trained first BERT-based neural network model to serve as an output layer to obtain a trained first BERT-based neural network model; the first Backpropagation-based optimization submodule is used for optimizing game element identification parameters of each layer of the trained first BERT-based neural network model based on Backpropagation according to standard game element marks corresponding to all game element training eigenvectors; and the game element judgment model determining submodule is used for determining the trained first BERT-based neural network model optimized based on Backpropagation as a game element judgment model.
7. The apparatus of claim 5, wherein the game play element contact neural network generating unit comprises: the second sub-word module, the second eigenvector construction sub-module, the second pre-training sub-module, the second adding sub-module, the second Backpropagation optimization sub-module and the game element contact neural network determination sub-module; the second participle submodule is used for participling the game chatting room text data game element association training corpus to obtain at least one game element association training vocabulary element; the second eigenvector construction sub-module is used for constructing corresponding game element contact training eigenvectors for all the game element contact training vocabulary elements; the second pre-training sub-module is used for taking all game element association training eigenvectors as input data of a second BERT-based neural network model for identifying game element association and performing layer-by-layer unsupervised pre-training on the second BERT-based neural network model, and the second BERT-based neural network model is formed by overlapping at least one multi-head attention mechanism layer; the second adding submodule is used for adding a neural network layer as an output layer in the pre-trained second BERT-based neural network model to obtain a second BERT-based neural network model; the second Backpropagation optimization submodule is used for optimizing game element connection judgment parameters of each layer of the second BERT-based neural network model based on Backpropagation according to standard game element connection marks among all game element connection training eigenvectors; and the game element contact neural network determining submodule is used for determining the second BERT-based neural network model optimized based on the Backpropagation as a game element contact neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211467546.9A CN115510857A (en) | 2022-11-22 | 2022-11-22 | Game education knowledge graph construction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211467546.9A CN115510857A (en) | 2022-11-22 | 2022-11-22 | Game education knowledge graph construction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115510857A true CN115510857A (en) | 2022-12-23 |
Family
ID=84514473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211467546.9A Pending CN115510857A (en) | 2022-11-22 | 2022-11-22 | Game education knowledge graph construction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115510857A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106934032A (en) * | 2017-03-14 | 2017-07-07 | 软通动力信息技术（集团）有限公司 | A kind of city knowledge mapping construction method and device |
| CN111460820A (en) * | 2020-03-06 | 2020-07-28 | 中国科学院信息工程研究所 | Network space security domain named entity recognition method and device based on pre-training model BERT |
| CN113609305A (en) * | 2021-07-27 | 2021-11-05 | 三峡大学 | Method and system for building geographical knowledge graph of film and television works based on BERT |
| CN114860946A (en) * | 2022-02-28 | 2022-08-05 | 广州启生信息技术有限公司 | Method and device for generating map network |
-
2022
- 2022-11-22 CN CN202211467546.9A patent/CN115510857A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106934032A (en) * | 2017-03-14 | 2017-07-07 | 软通动力信息技术（集团）有限公司 | A kind of city knowledge mapping construction method and device |
| CN111460820A (en) * | 2020-03-06 | 2020-07-28 | 中国科学院信息工程研究所 | Network space security domain named entity recognition method and device based on pre-training model BERT |
| CN113609305A (en) * | 2021-07-27 | 2021-11-05 | 三峡大学 | Method and system for building geographical knowledge graph of film and television works based on BERT |
| CN114860946A (en) * | 2022-02-28 | 2022-08-05 | 广州启生信息技术有限公司 | Method and device for generating map network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110795543B (en) | Unstructured data extraction method, device and storage medium based on deep learning | |
| JP6793975B2 (en) | Video-based Jobs Job Matching Servers and methods and computer-readable recording media containing programs to perform those methods | |
| CN103761975B (en) | Method and device for oral evaluation | |
| CN112487139B (en) | Text-based automatic question setting method and device and computer equipment | |
| CN108763326A (en) | A kind of sentiment analysis model building method of the diversified convolutional neural networks of feature based | |
| CN107992195A (en) | A kind of processing method of the content of courses, device, server and storage medium | |
| CN112599117B (en) | Model training method, model training device, voice recognition method, voice recognition device, electronic equipment and storage medium | |
| CN110321564B (en) | Multi-round dialogue intention recognition method | |
| CN111144097B (en) | Modeling method and device for emotion tendency classification model of dialogue text | |
| CN112329767A (en) | Contract text image key information extraction system and method based on joint pre-training | |
| CN107293290A (en) | The method and apparatus for setting up Speech acoustics model | |
| CN111898810B (en) | User loss prediction system based on teacher-student communication | |
| CN116010569A (en) | Online answering method, system, electronic equipment and storage medium | |
| CN116797417A (en) | Intelligent auxiliary system based on large language model | |
| CN113011196B (en) | Concept-enhanced representation and one-way attention-containing subjective question automatic scoring neural network model | |
| CN114360584A (en) | Phoneme-level-based speech emotion layered recognition method and system | |
| Yu et al. | The BURCHAK corpus: A challenge data set for interactive learning of visually grounded word meanings | |
| CN112749257A (en) | Intelligent marking system based on machine learning algorithm | |
| CN115510857A (en) | Game education knowledge graph construction method and device | |
| Shao et al. | Attention GAN-based method for designing intelligent making system | |
| CN115905513A (en) | Dialog summarization method based on denoising type question answering | |
| CN113963306B (en) | Courseware title making method and device based on artificial intelligence | |
| CN116257618A (en) | Multi-source intelligent travel recommendation method based on fine granularity emotion analysis | |
| CN113361615B (en) | Text classification method based on semantic relevance | |
| CN112927109B (en) | Teaching system of intelligent auxiliary practice practical training course category |
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: 20221223 |
|
| RJ01 | Rejection of invention patent application after publication |