An AI-based interactive teaching method, device and electronic device

Abstract

The present invention provides an artificial intelligence-based interactive teaching method, device and electronic device. The method includes: inputting the data of a strategy game into a pre-trained artificial intelligence model corresponding to the current student to output a first competitive strategy, generating a second competitive strategy through the input instructions of the current student for the strategy game, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current student; inputting the data of the strategy game into a pre-trained artificial intelligence model corresponding to other students to output a third competitive strategy, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current student; adjusting the curriculum syllabus and learning plan according to the first score, the second score and the data of the strategy game, and guiding the current student to enter the course learning to complete the teaching task in a targeted manner.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular, to an interactive teaching method, device, and electronic device based on artificial intelligence. Background Art

[0002] In the past decade, significant progress has been made in artificial intelligence technology. The general public, especially teenagers, have a strong demand for learning artificial intelligence knowledge. Due to the computer nature of artificial intelligence technology itself, the rapid development of artificial intelligence has given rise to numerous artificial intelligence education platforms. These platforms usually regard artificial intelligence as an additional content of programming teaching, combine it with programming courses, and guide students to learn the concept of artificial intelligence through programming. For example, low-code programming platforms such as Scratch enable students to learn artificial intelligence through programming by means of drag-and-drop programming. However, this educational model often confines artificial intelligence education to the programming language teaching stage, requiring learners to first master programming languages (mainly Python) before they can access artificial intelligence content. This not only has a relatively high threshold but also is difficult to attract teenagers, especially primary school students, because it is difficult for them to persist in long-term programming learning. In addition, this educational model overemphasizes algorithm learning while ignoring the importance of data in artificial intelligence, resulting in teenagers failing to fully understand that artificial intelligence is a data-based science.

[0003] How to combine artificial intelligence with education to enable students with different levels of mastery of artificial intelligence to take different courses is a problem that current educational platforms need to solve. Summary of the Invention

[0004] The present invention provides an interactive teaching method, device, and electronic device based on artificial intelligence to solve the defect that the existing teaching platforms cannot conduct different course learning according to the mastery levels of artificial intelligence of different students.

[0005] The present invention provides an interactive teaching method based on artificial intelligence for a teaching system, and the method includes:

[0006] Inputting the data of a strategy game into a pre-trained artificial intelligence model corresponding to the current student to output a first competitive strategy, generating a second competitive strategy through the input instructions of the current student for the strategy game, and then respectively controlling the two sides of the roles to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current student;

[0007] By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to other students, a third competitive strategy is output, and then based on the first competitive strategy and the third competitive strategy, the characters on both sides are respectively controlled to conduct a strategy game battle to obtain the second score of the current student;

[0008] Adjust the curriculum syllabus and learning plan according to the first score, the second score and the data of the strategy game, and guide the current student to enter the course study.

[0009] According to an interactive teaching method based on artificial intelligence provided by the present invention, the data of the strategy game includes the number of props and the positions of the props;

[0010] By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, a first competitive strategy is output, including:

[0011] According to the number of props and the positions of the props in the strategy game, multiple competitive strategies for obtaining props are generated; wherein, each competitive strategy includes the competition order and path of multiple props;

[0012] The multiple competitive strategies are input into the artificial intelligence model corresponding to the current student for calculation, the winning probability corresponding to each competitive strategy is obtained, and the competitive strategy with the highest winning probability is output as the first competitive strategy.

[0013] According to an interactive teaching method based on artificial intelligence provided by the present invention, a second competitive strategy is generated through the input instruction of the current student, including: determining the competition order and path of the prop through the input instruction of the prop displayed on the screen by the current student; generating a second competitive strategy according to the competition order and path of the prop.

[0014] According to an interactive teaching method based on artificial intelligence provided by the present invention, the data of the strategy game includes the number of props and the positions of the props;

[0015] By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to other students, a third competitive strategy is output, including:

[0016] According to the number of props and the positions of the props in the strategy game, multiple competitive strategies for obtaining props are generated; wherein, each competitive strategy includes the competition order and path of multiple props;

[0017] The multiple competitive strategies are input into the artificial intelligence model corresponding to other students for calculation, the winning probability corresponding to each competitive strategy is obtained, and the competitive strategy with the highest winning probability is output as the third competitive strategy.

[0018] An interactive teaching method based on artificial intelligence provided by the present invention, the artificial intelligence model corresponding to the current trainee is trained in the following ways:

[0019] Training data screening: Real-time capture and record the strategies and their results adopted by the trainee after the game round, store these data in a secure data storage module, and automatically select the most recent N winning games as winning strategy data;

[0020] Training data augmentation: Adopt a data augmentation algorithm to automatically generate more sample data from the winning strategy data, including winning sample data and losing sample data;

[0021] Feature engineering: Calculate the differences between the coordinate values of game props to generate new data points to expand the feature dimension of the dataset;

[0022] Algorithm selection: According to the conditions of the trainee, select an artificial intelligence algorithm that matches the trainee;

[0023] Parameter adjustment: According to the type of the selected algorithm, adjust the algorithm parameters;

[0024] Model establishment and optimization: Use the screened training data, the selected algorithm and parameters to train to obtain an artificial intelligence model, and save it to the operation and maintenance module.

[0025] According to an interactive teaching method based on artificial intelligence provided by the present invention, respectively control the two sides of the role to carry out a strategic game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current trainee, including:

[0026] Based on the first competitive strategy and the second competitive strategy, respectively control the two sides of the role to obtain props according to the competition order and path of their respective props;

[0027] If one of the roles obtains any prop first, then the prop is removed from the competition order and path of the two roles, and their respective competition order and path are updated;

[0028] Continue to control the two sides of the role to continue to obtain props according to their respective updated competition order and path of the props until all the props are obtained, and count the number of props obtained by each of the two sides of the role;

[0029] Determine the first score of the current trainee according to the difference in the number of props.

[0030] According to an interactive teaching method based on artificial intelligence provided by the present invention, respectively control the two sides of the role to carry out a strategic game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current trainee, specifically including:

[0031] Based on the first competition strategy and the third competition strategy, control the characters on both sides to obtain items according to the competition order and path of their respective items;

[0032] If one of the characters obtains any item first, eliminate that item from the competition order and path of both characters, and update their respective competition order and path;

[0033] Continue to control the characters on both sides to continue obtaining items according to their respective updated competition order and path of items until all items are obtained, and count the number of items obtained by each character;

[0034] Calculate based on the initial score of the current student's character, the initial scores of other students' characters, the number of items obtained by the current student, and the number of items obtained by other students to determine the second score of the current student.

[0035] According to an interactive teaching method based on artificial intelligence provided by the present invention, adjust the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guide the current student to enter the course study, including:

[0036] If the score obtained by the current student in the first competition strategy exceeds the first set score threshold, unlock the competition mode of controlling the characters on both sides to conduct a strategy game battle based on the first competition strategy and the second competition strategy, and add basic artificial intelligence popular science teaching content to the curriculum syllabus of the current student;

[0037] In the case where the first score of the current student exceeds the second set score threshold, add an explanation of the basic principles of artificial intelligence to the curriculum syllabus of the current student;

[0038] In the case where the second score of the current student exceeds the third set score threshold, continue to add relevant knowledge of building an artificial intelligence model to the curriculum syllabus of the current student.

[0039] The present invention also provides an interactive teaching device based on artificial intelligence, including the following modules:

[0040] A human-computer battle module, which is used to input the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, output the first competition strategy, generate the second competition strategy through the input instructions of the current student for the strategy game, and then control the characters on both sides to conduct a strategy game battle based on the first competition strategy and the second competition strategy to obtain the first score of the current student;

[0041] The machine battle module is used to input the data of the strategy game into the pre-trained artificial intelligence model corresponding to other students, output the third competition strategy, and then control the characters on both sides to conduct a strategy game battle based on the first competition strategy and the third competition strategy respectively, so as to obtain the second score of the current student;

[0042] The course selection module is used to adjust the curriculum syllabus and learning plan according to the first score, the second score and the data of the strategy game, and guide the current student to enter the course study.

[0043] The present invention also provides an electronic device, including a memory, a processor and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the artificial intelligence-based interactive teaching method as described in any one of the above.

[0044] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the artificial intelligence-based interactive teaching method as described in any one of the above.

[0045] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the artificial intelligence-based interactive teaching method as described in any one of the above.

[0046] The artificial intelligence-based interactive teaching method and device provided by the present invention respectively obtain the first score by having the second competition strategy input by the current student fight against the first competition strategy output by the pre-trained artificial intelligence model, and obtain the second score by having the first competition strategy output by the artificial intelligence model corresponding to the current student fight against the third competition strategy output by the artificial intelligence model corresponding to other students. Furthermore, the total score of the current student can be determined to represent the degree of mastery of the current student in artificial intelligence. Then, according to the first score, the second score and the data of the strategy game, the curriculum syllabus and learning plan are adjusted, and the current student is guided to enter the course study, so as to realize the learning of corresponding courses for students with different degrees of mastery of artificial intelligence in the way of game battles between students and artificial intelligence models, and thus complete the teaching task in a targeted manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0048] Figure 1It is a schematic structural diagram of the interactive teaching system provided by the present invention.

[0049] Figure 2 It is one of the schematic flow diagrams of the artificial intelligence-based interactive teaching method provided by the present invention.

[0050] Figure 3 It is a schematic interface diagram of a strategy game provided by the present invention.

[0051] Figure 4 It is the second schematic flow diagram of the artificial intelligence-based interactive teaching method provided by the present invention.

[0052] Figure 5 It is a schematic diagram of the battle process of a strategy game provided by the present invention.

[0053] Figure 6 It is the third schematic flow diagram of the artificial intelligence-based interactive teaching method provided by the present invention.

[0054] Figure 7 It is a schematic structural diagram of the artificial intelligence-based interactive teaching device provided by the present invention.

[0055] Figure 8 It is a schematic structural diagram of an electronic device provided by the present invention. Detailed implementation manners

[0056] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] The following is combined with Figures 1-7 to describe the artificial intelligence-based interactive teaching method and device of the present invention. The teaching method of this embodiment is used for an interactive teaching system.

[0058] First, a schematic description of each functional module of the interactive teaching system is given. Refer to Figure 1 , the interactive teaching system includes:

[0059] 1. Front-end interaction module.

[0060] Login module: Students log in to the system through the login module.

[0061] Winning game screening module: Students can browse previous game results and decide which game data can be used for the training of their own artificial intelligence models.

[0062] Algorithm Selection Module: Learners can build an artificial intelligence model for themselves to adjust parameters and algorithms. Learners can select the algorithms they want to use from a preset algorithm library and adjust the corresponding parameters of the algorithms.

[0063] Mode Selection Module: Learners select the game mode in this module, which can be divided into three modes according to the design purpose:

[0064] 1) Beginner Mode: In this mode, learners play against an artificial intelligence model using a random strategy, learn basic operations, and collect initial game data.

[0065] 2) Advanced Mode: In this mode, the system will call the fully automatic model building function in the artificial intelligence model building module and use the game data generated by learners in the beginner mode to train an artificial intelligence model based on the learners' own competitive strategies. Learners will play against the game characters driven by this model. After learners achieve a certain number of victories (e.g., 10 games), the system will retrain the model using the same method and let learners play against the characters driven by the newly trained artificial intelligence model in a new game.

[0066] 3) AI Battle Mode: Learners will use the model created in the artificial intelligence model building module to control their own characters and play against the characters controlled by the artificial intelligence models created by other learners through the matching module in the battle module. After the battle, the models of both sides will be adjusted in score through the scoring module, and finally, the rankings on the leaderboard will be updated through the ranking module.

[0067] 2. Data Storage Module: It includes a database and an artificial intelligence model operation and maintenance module.

[0068] Database: Store all users' login information, game strategies and win-loss information, activity information organized by the activity organization module, and all generated digital information through a database based on Structured Query Language (SQL).

[0069] Artificial Intelligence Model Operation and Maintenance Module: This module is used to store and manage various artificial intelligence models established by users and the system through the artificial intelligence model building module.

[0070] 3. Artificial Intelligence Model Building Module:

[0071] 1) Data Screening:

[0072] After the game round is completed, the system will capture and record in real-time the strategies adopted by learners and their game results. These data will be stored in a secure module designed specifically for data storage to ensure data integrity and accessibility.

[0073] The system has two methods for screening strategy data:

[0074] The system will automatically select the most recent 10 winning games as strategy data. With the participation of the trainees, the system will guide the trainees to manually screen the key strategy data in the historical winning games through an intuitive user interface. The system will adopt a data augmentation algorithm to automatically generate more winning samples and corresponding losing samples from the existing winning game data to expand the scale and diversity of the dataset. This step is crucial for improving the generalization ability and robustness of the model. The specific implementation steps of the algorithm are as follows:

[0075] a) The system will import the user's strategy information, including the number of items, the specific strategy for the user to obtain items, and the acquisition results. The system will screen the items successfully obtained by the user and import their coordinate data.

[0076] b) For the imported coordinate data, the system will perform the following processing on the x coordinate (denoted as P x ) and the y coordinate (denoted as P y ) to generate new coordinate data: P x ' and P y ':

[0077] P x ' = P x + max(0,min(255, P x +random_sign)

[0078] P y ' = P y + max(0,min(255, P y +random_sign)

[0079] where random_sign is a randomly selected sign, which can be 1 (indicating an increase of 1) or -1 (indicating a decrease of 1).

[0080] c) The system will use the newly generated coordinate data P x ′ and 𝑃 y ′ to replace the original x coordinate (P x value) and y coordinate (P y value) data, and add these data as a new winning game dataset to the dataset screened above.

[0081] In addition, the system will perform the following steps to apply the feature engineering algorithm:

[0082] a) For each user data in the dataset, the system will calculate the x coordinate value of the item (𝑃𝑥 ) and the difference between the y - coordinate value (𝑃 𝑦 ).

[0083] b) Taking a dataset containing 7 sets of competitive items as an example, the system will subtract the x - coordinate values (from 𝑃 𝑥1 to 𝑃 𝑥7 ) of each item pairwise to generate new data points. Specifically, for any two coordinate values 𝑃 𝑥𝑖 and 𝑃 𝑥𝑗 (where 𝑖≠𝑗), calculate the difference, and the formula is: 𝐷 𝑥𝑖𝑗 = 𝑃 𝑥𝑖 −𝑃 𝑥𝑗 .

[0084] c) These calculated differences 𝐷 𝑥𝑖𝑗 will be used as new data points and added to the original dataset.

[0085] Through this process, the system can expand the feature dimension of the dataset and provide additional information for subsequent data analysis and model training.

[0086] 2) Algorithm selection: The system will select and adjust the artificial intelligence algorithm through a series of detailed steps. The specific implementation order is as follows:

[0087] a) The system will consider the personalized selection of the trainee. If the trainee has specified a preferred model, the system will adopt this selection.

[0088] b) The system will evaluate the number of items in the dataset. If the number exceeds a specific value (e.g., 10), the system will select a neural network model because neural networks are suitable for processing datasets with a large number of features.

[0089] c) The system will analyze the winning percentage in the enhanced dataset. If the winning percentage is lower than a specific threshold, such as 10%, the system will select the XgBoost model, which is suitable for processing imbalanced datasets; if the winning percentage is higher than another threshold, such as 80%, then select the random forest model to take advantage of its superiority in processing diverse datasets.

[0090] d) If none of the above conditions are met, the system will default to selecting the logistic regression model.

[0091] 3) Parameter selection: After selecting the model, the system will perform intelligent parameter adjustment. The system will select specific parameters for adjustment according to the algorithm type selected by the trainee or the user himself / herself as described above. The specific content is as follows:

[0092] a) Adjust the number of layers, the number of neurons, the learning rate, and the regularization parameter of the neural network;

[0093] b) Set the depth of the tree, learning rate, number of trees, and regularization parameters of XgBoost;

[0094] c) Optimize the number of trees, minimum number of samples for splitting, and feature selection method of the random forest;

[0095] d) Adjust the regularization strength and number of iterations of logistic regression.

[0096] 4) Establish a model: The system will automatically use the screened data, selected algorithms, and parameters to train an artificial intelligence model and save it to the artificial intelligence model operation and maintenance module. During the model training process, the system implements cross-validation to evaluate the model performance under different parameter settings, ensuring that the selected model and parameter configuration can provide high prediction accuracy and generalization ability. At the same time, the system will continuously monitor the key performance indicators of the model, such as accuracy, recall rate, and F1 score, to ensure the effectiveness of training. According to the results of cross-validation and performance monitoring data, the system will adjust the parameters of the model, including learning rate, regularization parameters, depth of the tree, etc., to optimize the training effect.

[0097] After completing these steps, the system will save the final artificial intelligence model to the operation and maintenance module, providing the most suitable algorithm and fine-tuning for each student and dataset to ensure the best training effect.

[0098] 4. Battle module: This module consists of a scoring module, a matching module, and a leaderboard module.

[0099] 1) Scoring module: After a game ends, calculate the latest score of the user role through a specific calculation method and store it in the database.

[0100] 2) Matching module: When a student enters the AI battle mode, the system will automatically start the matching function. The selection process is as follows: From the leaderboard, the system will select 20 students whose rankings are similar to the student's, including 10 students ranked before the student and 10 students ranked after the student. Then, randomly select a student from these 20 students, and then the system will retrieve the latest trained artificial intelligence model of this student for battle.

[0101] 3) Leaderboard module: Rank all students or students within a specific activity by retrieving data from the database module. The leaderboard can be viewed on the front-end interaction interface.

[0102] 5. Activity organization module: Includes an information module and a management module.

[0103] The information module will display the basic information of the activity, including but not limited to: activity organizer, teaching teacher, activity time, activity rewards, QR code for participating in the activity, etc.

[0104] Management module: Administrators and event supervisors can establish events through this module and manage the quantity of events and start / end times.

[0105] Figure 2 It is one of the schematic flowcharts of the interactive teaching method based on artificial intelligence provided by the present invention. This method is used for the above teaching system and includes:

[0106] Step 201: Input the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, output the first competitive strategy, generate the second competitive strategy through the input instructions of the current student for the strategy game, and then control the two sides of the role to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy respectively to obtain the first score of the current student.

[0107] In this embodiment, the strategy game refers to a game in which both sides of the battle formulate competitive strategies and then control the roles to conduct a battle based on the formulated competitive strategies. The two sides of the battle can be student-student, student-machine (i.e., the artificial intelligence module), and machine-machine.

[0108] See Figure 3 , Figure 3 shown is a schematic diagram of an interface of the strategy game described in this embodiment.

[0109] The operation interface provides an interactive page for the student to operate. As can be seen from Figure 3 , the page is mainly composed of a chessboard. Role 1 (controlled by the student or the artificial intelligence model generated by the student) appears at one end of the square game interface (such as the upper left corner or the lower left corner), and at the same time, Role 2 (automatically generated or the artificial intelligence model generated by other students) appears at the other end of the diagonal of the square game interface. The student selects strategies by finger touch (mobile platform) or mouse selection (such as Mac, Linux, and PC platforms) with the goal of obtaining more props.

[0110] Among them, the artificial intelligence model corresponding to the current student is trained in the following manner:

[0111] The system will follow the following steps to ensure the establishment and optimization of the artificial intelligence model:

[0112] Training data screening: The system will capture and record in real time the strategies adopted by the student after the game round and their results, and store these data in a secure data storage module. The system will automatically select the most recent N winning rounds as the winning strategy data, or with the participation of the student, manually screen the key strategy data through the user interface. Among them, N can be 10, 15, 20, etc.

[0113] Data Augmentation: The system will adopt data augmentation algorithms to automatically generate more sample data from the winning game data, including winning sample data and losing sample data, to expand the scale and diversity of the dataset. The specific steps include: importing user strategy information, screening the successfully obtained prop coordinate data, and processing it to generate new coordinate data, which will be used as part of the winning game dataset.

[0114] Feature Engineering: The system will calculate the differences between prop coordinate values to generate new data points, thereby expanding the feature dimensions of the dataset. For example, for a dataset of 7 sets of competitive props, the system will calculate the differences between the x coordinate values of every two props and add these differences as new data points to the dataset.

[0115] Algorithm Selection: The system will select an artificial intelligence algorithm that matches the trainee according to conditions such as the trainee's personalized choice, the number of props, and the winning game ratio. If the trainee has a preference, the system will give priority to it; if the number of props exceeds 10, the neural network model will be selected; when the winning game ratio is lower than 10%, the XgBoost model will be selected, and when it is higher than 80%, the random forest model will be selected; in other cases, the logistic regression model will be selected by default.

[0116] Parameter Tuning: The system will adjust the algorithm parameters according to the selected algorithm type. This includes the number of layers, the number of neurons, the learning rate, and the regularization parameter of the neural network; the tree depth, the learning rate, the number of trees, and the regularization parameter of XgBoost; the number of trees, the minimum number of samples for splitting, and the feature selection method of the random forest; the regularization strength and the number of iterations of the logistic regression.

[0117] Model Building and Optimization: The system will automatically use the screened data, the selected algorithm, and parameters to train an artificial intelligence model and save it to the operation and maintenance module. During the training process, the system implements cross-validation to evaluate the model performance under different parameter settings, continuously monitors performance metrics such as accuracy, recall, and F1 score, and intelligently adjusts the model parameters to optimize the training effect. Finally, the system will save the optimized artificial intelligence model to the operation and maintenance module to ensure the most suitable algorithm and fine-tuning for each trainee and dataset, achieving the optimal training effect.

[0118] Through the above steps:

[0119] First, after the trainee plays the game using the first or second competitive strategy, or the competitive strategy selected according to personal preference, the results generated will be automatically recorded and stored in a dedicated data storage module by the system. Subsequently, under the guidance of the system, the trainee can manually screen out the winning game strategy data from the historical data, and these data will be used as the basis for the training data.

[0120] The system will automatically utilize the win data and generate additional win data and their corresponding loss data through a data augmentation algorithm. Meanwhile, the system will apply feature engineering algorithms to expand the features in the dataset to enhance the richness and predictive ability of the dataset. These carefully processed data will be integrated to form a comprehensive training dataset.

[0121] Based on this training dataset, the system will apply a preset algorithm framework, combined with the inherent features of the dataset, such as the quantity of win data, the diversity of props, and the personalized choices of trainees, to recommend or select the most suitable artificial intelligence algorithms for users. These algorithms include, but are not limited to, logistic regression, XGBoost, random forest, and neural network, etc., aiming to assist users in efficiently training an outstanding artificial intelligence model.

[0122] Step 201 is in the form of a human-machine battle. The trainee battles with the artificial intelligence model trained by himself / herself, and then the first score is counted. After the trainee achieves a certain number of victories (for example, 10 victories), the system will retrain the model using the same method and let the trainee battle with the character driven by the newly trained artificial intelligence model in a new game.

[0123] Step 202: Input the data of the strategy game into the pre-trained artificial intelligence model corresponding to other trainees, output the third competitive strategy, and then respectively control the characters of both sides to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current trainee.

[0124] Step 202 is in the form of a machine-machine battle. By replacing the trainees with pre-trained artificial intelligence models among the trainees, the characters of both sides are controlled to conduct a strategy game battle to obtain the second score of the trainee.

[0125] From the leaderboard, the system will select 20 users whose rankings are similar to the user's, including 10 users ranked before the user and 10 users ranked after the user. Then, a user is randomly selected from these 20 users, and the system will retrieve the artificial intelligence model newly trained by this user for a battle.

[0126] Step 203: Adjust the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guide the current trainee to enter the course study.

[0127] The following is the operation process of this method:

[0128] 1) The system first prepares a blank curriculum syllabus for each student. When the score obtained by the student in the first competitive strategy exceeds the first set score threshold (e.g., 10 points), the competitive mode of controlling the strategy game battles of both sides' characters based on the first competitive strategy and the second competitive strategy is unlocked, and basic artificial intelligence popular science teaching content is added to the curriculum syllabus of the current student. This content mainly involves the applications of artificial intelligence and programming in daily life and future development trends. Once unlocked, the scores continuously obtained by the student in the first competitive strategy will no longer increase the teaching content.

[0129] 2) In the case where the first score of the current student exceeds the second set score threshold (e.g., 100 points), the system will automatically add an explanation of the basic principles of artificial intelligence to the student's curriculum syllabus.

[0130] 3) In the case where the second score of the current student exceeds the third set score threshold (e.g., 200 points), the system will continue to add curriculum content to the student's curriculum syllabus, including knowledge related to the basics of Python programming and the establishment of artificial intelligence models.

[0131] 4) Additionally, in the case where the second score of the current student exceeds the third set score threshold (e.g., 200 points), the system will automatically detect the student's performance in the game. After meeting specific conditions, the system will continue to add more teaching content to the student's curriculum syllabus, specifically including:

[0132] a) If the student's winning rate exceeds 80%, or the second game score is obtained by using more than three artificial intelligence models, the system will add more in-depth knowledge of artificial intelligence and Python programming.

[0133] b) If the student uses all the preset artificial intelligence models to participate in the game and obtains the second game score, the system will unlock all the curriculum content.

[0134] c) After the student reaches a very high score (e.g., 2000 points), finally, the curriculum content of building an artificial intelligence product and Python algorithms from scratch will be unlocked for the student.

[0135] Meanwhile, the system allows the student to select and enter the course study at any time. The system will merge the course content with high similarity according to the automatically generated teaching syllabus and the teaching content already completed by the student to generate a new teaching syllabus.

[0136] The system will automatically match the educational resources (such as videos, PPTs, and text materials) in the resource library according to the content of the teaching syllabus and attach links to the teaching syllabus. The specific matching method is as follows:

[0137] a) Use the pre-trained BERT model to vectorize the entry content in the teaching syllabus to generate a digital matrix M A。

[0138] b) Similarly, vectorize the introduction content of all courses in the resource library to generate a series of digital matrices M Bi , where i is the course serial number.

[0139] c) Calculate the cosine similarity between each vector in matrix M A and M B , and select the course corresponding to the maximum value as the content that best matches the corresponding course syllabus.

[0140] d) Convert the entries in the course syllabus into hyperlinks that directly link to the course content.

[0141] In addition, automatically attach a learning date to the course syllabus, and the date is formulated according to the expected completion time.

[0142] The adjusted syllabus will help trainees immediately start learning knowledge related to artificial intelligence, improve learning efficiency, ensure the personalization and pertinence of teaching content, and meet the learning needs and progress of different trainees.

[0143] The interactive teaching method based on artificial intelligence provided by the present invention respectively obtains a first score by having the second competitive strategy input by the current trainee fight against the first competitive strategy output by the pre-trained artificial intelligence model, and obtains a second score by having the first competitive strategy output by the artificial intelligence model corresponding to the current trainee fight against the third competitive strategy output by the artificial intelligence model corresponding to other trainees, and then can determine the total score of the current trainee to represent the degree of mastery of artificial intelligence by the current trainee. Furthermore, adjust the course syllabus and learning plan according to the first score, the second score and the data of the strategy game, and guide the current trainee into the course learning, so as to complete the teaching task in a targeted manner.

[0144] The invention mainly improves and optimizes the concept and architecture parts of the traditional artificial intelligence-based game, specifically manifested in:

[0145] 1) Establish and design a no-code artificial intelligence model establishment system to help trainees establish their own artificial intelligence models using the data they generate and manage these artificial intelligence models.

[0146] 2) Under this no-code system, the present invention creates a platform that allows trainees to fight against an artificial intelligence model that continuously learns the data it generates.

[0147] 3) This example provides a platform for trainees to establish their own artificial intelligence models and automatically compete and score with the artificial intelligence models of other trainees.

[0148] Further, the data of the strategy game includes the number of items and the positions of the items; in step 201, by inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, a first competitive strategy is output, including: generating multiple competitive strategies for obtaining items according to the number of items and the positions of the items in the strategy game; wherein each competitive strategy includes the competition order and path of multiple items; inputting the multiple competitive strategies into the artificial intelligence model corresponding to the current student for calculation, obtaining the winning probability corresponding to each competitive strategy, and outputting the competitive strategy with the highest winning probability as the first competitive strategy.

[0149] Further, in step 201, a second competitive strategy is generated through the input instruction of the current student, including: determining the competition order and path of the item through the input instruction of the item displayed on the screen by the current student; generating a second competitive strategy according to the competition order and path of the item.

[0150] Further, step 202 specifically includes: generating multiple competitive strategies for obtaining items according to the number of items and the positions of the items in the strategy game; wherein each competitive strategy includes the competition order and path of multiple items; inputting the multiple competitive strategies into the artificial intelligence model corresponding to other students for calculation, obtaining the winning probability corresponding to each competitive strategy, and outputting the competitive strategy with the highest winning probability as the third competitive strategy.

[0151] Wherein, in this embodiment, the system will call the model pre-saved by the student in the artificial intelligence model storage module and use the competitive strategy output by it as the third competitive strategy.

[0152] Further, referring to Figure 4 , in step 201, based on the first competitive strategy and the second competitive strategy, the two roles are respectively controlled to conduct a strategy game battle, and the first score of the current student is obtained, including the following steps 401 to 404:

[0153] 401. Based on the first competitive strategy and the second competitive strategy, control the two roles to obtain items according to the competition order and path of their respective items.

[0154] 402. If one of the roles obtains any item first, the item is removed from the competition order and path of the two roles, and their respective competition order and path are updated.

[0155] 403. Continue to control the two roles to obtain items according to their respective updated competition order and path of the items until all items are obtained, and count the number of items obtained by each of the two roles.

[0156] 404. Determine the first score of the current student according to the difference in the number of items.

[0157] Specifically, the calculation of the first score is shown in the following formula:

[0158] S r = N a - N b

[0159] Where: S r is the change in the user's score, N a is the number of items obtained by the user's role, and N b is the number of items obtained by the opponent's role.

[0160] Example: If the current student obtains 5 items and the opponent obtains 2 items, the user's score increases by 3; if the current student obtains 4 items and the opponent obtains 5 items, the user's score decreases by 1.

[0161] See Figure 5 , Figure 5 which shows the battle process of a strategy game.

[0162] Interface generation: Character 1 is controlled by the student and appears at one end of the square game interface (e.g., the upper left corner or the lower left corner), while Character 2 is controlled by the artificial intelligence model generated by the student and appears at the other end of the diagonal of the square game interface. An odd number of competing items (e.g., 7) appear on the interface, as Figure 5 shown in a.

[0163] Before the game battle, a fairness check is first performed: The system will try different initial food distributions until it finds a distribution plan such that the sum of the coordinate values (x and y values) of the initial items distributed on the chessboard is within an acceptable tolerance range compared to the theoretically average distributed item coordinate values (calculation method: (chessboard side length - 1) × number of items). Once such a distribution is found, the function will return this distribution plan and end. The tolerance calculation example:

[0164] Lower limit: (chessboard side length - 1) × number of items × 75%;

[0165] Upper limit: (chessboard side length - 1) × number of items × 125%;

[0166] User strategy selection: The user selects the competing items they want to obtain in sequence by using the mouse or by touching the screen of a mobile device. The acquisition order and the preset route will be presented on the screen, as Figure 5 shown in b.

[0167] Opponent's strategy selection: The system loads the corresponding model from the artificial intelligence model operation and maintenance module (the artificial intelligence model established by the user's own data). The system exhaustively generates all strategies for obtaining props according to the number (N) and positions of the props (a total of N! groups of strategies). Then, it calculates the winning probabilities corresponding to all N! groups of strategies through the loaded model, and selects the group with the highest winning rate as the output strategy.

[0168] Method for information synchronization and update: After the trainee confirms the strategy and selects "Start the game", the two characters will immediately act automatically and move towards the next obtainable prop according to the established strategy. The two characters compete for the acquisition of props according to the preset strategy. All information will be summarized and calculated in the cloud and displayed on the user side.

[0169] Information update: If any character obtains a new prop, the prop will be removed from the movement routes of both characters, and the movement routes of both characters will be updated in the computing center of the cloud. The first party to contact the competing prop in terms of time is regarded as the successful acquirer.

[0170] When both parties reach the prop at the same time: If both parties contact the competing prop at the same time, each party has a 50% probability of being determined to successfully acquire the prop.

[0171] Example: See Figure 5 c and Figure 5 d. For example, if the strategy set by the trainee is 1-2-3-4-5, and the strategy of the opponent character is generated as 2-3-4-5-1, and the trainee's character takes the lead in obtaining prop 1. Then the movement routes of both parties will be updated simultaneously. The movement route of the trainee's character is corrected to 2-3-4-5, and the movement route of the opponent character is corrected to 2-3-4-5. After that, the trainee's character will turn and start moving towards prop 2, while the movement route of the opponent character remains unchanged. At this time, the opponent character will give priority to obtaining prop 2. At this time, the strategies of the trainee and the opponent characters will be corrected to 3-4-5 at the same time, and their routes will be adjusted and they will move towards prop 3 simultaneously. The two parties will continue this process until all props are acquired.

[0172] After all props are acquired, the game will be settled according to the number of props obtained by both parties and the winner and loser will be determined. For example, if the user character obtains 3 props and the opponent character obtains 4 props, the user character is determined to be defeated. The game result and all data are recorded in the data storage unit.

[0173] Furthermore, see Figure 6 In step 202, respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current trainee, specifically including:

[0174] 601. Control the characters of both sides to obtain items according to the competition order and path of their respective items based on the first competition strategy and the third competition strategy respectively.

[0175] 602. If one of the characters obtains any item first, remove the item from the competition order and path of the characters of both sides, and update their respective competition orders and paths.

[0176] 603. Continue to control the characters of both sides to obtain items according to their respective updated competition orders and paths of items until all items are obtained, and count the number of items obtained by the characters of both sides respectively.

[0177] 604. Calculate based on the initial score of the current trainee's character, the initial scores of the characters of other trainees, the number of items obtained by the current trainee, and the number of items obtained by other trainees to determine the second score of the current trainee.

[0178] Among them, the calculation process of the second score is as follows:

[0179]

[0180]

[0181]

[0182] : Take 1 when winning on one's own side and 0 when losing.

[0183]

[0184] Among them: S r is the settlement score after the user's game, S a is the initial score of the user's character, S b is the initial score of the opponent's character, N a is the number of items obtained by the user's character, Q a and Q b are intermediate parameters in the calculation process; Ea is to calculate the score ratio of the current user (User A) and the other user (User B) using the amplified scores of both sides; K and c are both preset parameters, set to 16 and 200 respectively.

[0185] The finally output S r will be rounded. If S r is calculated to be negative after calculation, then return 0.

[0186] Example: At the beginning, the score of one's own side is 456, and the score of the opponent is 567. If one's own side obtains 4 props and wins, then the score after the game for one's own side is 472. At the beginning, the score of one's own side is 300, and the score of the opponent is 300. If one's own side obtains 3 props and loses, then the score after the game for one's own side is 295.

[0187] For the situation of controlling the characters of both sides to conduct a machine-to-machine strategy game battle based on the first competitive strategy and the third competitive strategy respectively, it is similar to the process of the man-machine battle described above Figure 5 and will not be elaborated here.

[0188] In the whole teaching process, the trainees first fight against computer characters that generate the second competitive strategy and the chance random strategy, accumulate certain experience and achieve certain victories, and then unlock and experience the advanced mode. In the advanced mode, the trainees will fight against characters driven by the first competitive strategy and the artificial intelligence model generated based on their own data. The winning rate of the trainees will drop sharply, and at the same time, it will stimulate the trainees to think about how to defeat this character based on the artificial intelligence model. Whether successful or not, interested trainees will be guided to build their own artificial intelligence models using their own data, and fight against the characters established by other users and participate in competitions.

[0189] The example of the present invention can serve as an experience course and a leading course for artificial intelligence teaching. During the lesson preparation process, teachers can arrange for trainees to experience the example of the present invention and feel the changes brought by artificial intelligence through actual operations. In this process, especially, trainees can experience the relationship between artificial intelligence and data by quickly building an artificial intelligence model, and quickly develop an interest in artificial intelligence. Since it is in the form of a strategy game and can be quickly experienced, relatively young trainees (aged 7 to 15) can also understand the corresponding content. This makes the example of the present invention significantly different from previous products as a teaching tool. The example of the present invention creates a brand-new AI experience environment that teenagers can understand, without requiring teenagers to understand topics such as housing prices, image recognition, or word processing that are difficult for teenagers to understand.

[0190] At the same time, the example of the invention provides teaching topics for subsequent artificial intelligence teaching. Teachers can design short and effective courses with this example of the invention as the theme. Through the study of the courses, trainees will improve their understanding of artificial intelligence, and then return to the example of the present invention to quickly build a new artificial intelligence model. This model will quickly bring feedback to the trainees through the competition, thereby helping them learn.

[0191] At the same time, one of the major learning challenges of artificial intelligence is dealing with a large amount of data. However, it is very difficult to obtain a large amount of data. In the embodiments of the present invention, a large amount of highly regular data will be generated through games. Through these regular data, the process for zero - based learners to get in touch with artificial intelligence from scratch can be greatly accelerated, and a foundation for future learning can be laid.

[0192] The following describes the interactive teaching device based on artificial intelligence provided by the present invention. The interactive teaching device based on artificial intelligence described below can be referred to correspondingly with the interactive teaching method based on artificial intelligence described above.

[0193] The present invention provides an interactive teaching device based on artificial intelligence. Refer to Figure 7 , including:

[0194] A human - machine battle module 701, configured to input the data of a strategy game into a pre - trained artificial intelligence model corresponding to the current learner, output a first competitive strategy, generate a second competitive strategy through the input instruction of the current learner for the strategy game, and then respectively control the two - side characters to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy, so as to obtain the first score of the current learner;

[0195] A machine - machine battle module 702, configured to input the data of a strategy game into a pre - trained artificial intelligence model corresponding to other learners, output a third competitive strategy, and then respectively control the two - side characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy, so as to obtain the second score of the current learner;

[0196] A course selection module 703, configured to adjust the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guide the current learner to enter course learning.

[0197] Optionally, the data of the strategy game includes the number of props and the positions of the props;

[0198] The human - machine battle module 701 is specifically configured to:

[0199] Generate multiple competitive strategies for obtaining props according to the number of props and the positions of the props in the strategy game; wherein, each competitive strategy includes the competition order and path of multiple props;

[0200] Input the multiple competitive strategies into the artificial intelligence model corresponding to the current learner for calculation, obtain the winning probability corresponding to each competitive strategy, and output the competitive strategy with the highest winning probability as the first competitive strategy.

[0201] Optionally, the human-machine battle module 701 is specifically configured to: determine the competition order and path of the props through the input commands of the current trainee for the props displayed on the screen; generate a second competition strategy according to the competition order and path of the props.

[0202] Optionally, the data of the strategy game includes the number of props and the positions of the props;

[0203] The machine battle module 702 is specifically configured to:

[0204] Generate multiple competition strategies for obtaining props according to the number of props and the positions of the props in the strategy game; wherein, each competition strategy includes the competition order and path of multiple props;

[0205] Input the multiple competition strategies into the artificial intelligence models corresponding to other trainees for calculation, obtain the winning probabilities corresponding to each competition strategy, and output the competition strategy with the highest winning probability as the third competition strategy.

[0206] Optionally, it further includes: a model training module, which is used to:

[0207] Training data screening: Real-time capture and record the strategies adopted by trainees after game matches and their results, store these data in a secure data storage module, and automatically select the most recent N winning matches as winning strategy data;

[0208] Training data enhancement: Use data enhancement algorithms to automatically generate more sample data from the winning strategy data, including winning sample data and losing sample data;

[0209] Feature engineering: Calculate the differences between the coordinate values of game props to generate new data points to expand the feature dimension of the dataset;

[0210] Algorithm selection: Select an artificial intelligence algorithm that matches the trainee according to the conditions of the trainee;

[0211] Parameter adjustment: Adjust the algorithm parameters according to the type of the selected algorithm;

[0212] Model establishment and optimization: Use the screened training data, the selected algorithm and parameters to train an artificial intelligence model, and save it to the operation and maintenance module.

[0213] Optionally, the human-machine battle module 701 is specifically configured to: respectively control the two sides of the role to obtain props according to the competition order and path of their respective props based on the first competition strategy and the second competition strategy;

[0214] If one of the role obtains any prop first, the competition order and path of both roles will exclude this prop, and update their respective competition order and path;

[0215] Continue to control the two characters to continue obtaining items according to the competition order and path of their respective updated items until all items are obtained, and count the number of items obtained by each of the two characters;

[0216] Determine the first score of the current student according to the difference in the number of items.

[0217] Optionally, the machine battle module 702 is specifically used for:

[0218] Based on the first competition strategy and the third competition strategy, respectively control the two characters to obtain items according to the competition order and path of their respective items;

[0219] If one of the characters obtains any item first, the item is removed from the competition order and path of the two characters, and their respective competition orders and paths are updated;

[0220] Continue to control the two characters to continue obtaining items according to the competition order and path of their respective updated items until all items are obtained, and count the number of items obtained by each of the two characters;

[0221] Calculate based on the initial score of the current student's character, the initial score of the other student's character, the number of items obtained by the current student, and the number of items obtained by the other student to determine the second score of the current student.

[0222] Optionally, the course selection module 703 is specifically used for:

[0223] If the score obtained by the current student in the first competition strategy exceeds the first set score threshold, unlock the competition mode that controls the two characters to conduct a strategy game battle based on the first competition strategy and the second competition strategy, and add basic artificial intelligence popular science teaching content to the current student's curriculum syllabus;

[0224] In the case where the first score of the current student exceeds the second set score threshold, add an explanation of the basic principles of artificial intelligence to the current student's curriculum syllabus;

[0225] In the case where the second score of the current student exceeds the third set score threshold, continue to add relevant knowledge about building an artificial intelligence model to the current student's curriculum syllabus.

[0226] Figure 8 Illustrates a schematic diagram of the physical structure of an electronic device, such as Figure 8As shown, the electronic device may include: a processor 810, a communications interface 820, a memory 830, and a communication bus 840. Among them, the processor 810, the communications interface 820, and the memory 830 complete communication with each other through the communication bus 840. The processor 810 may call logic instructions in the memory 830 to execute an interactive teaching method based on artificial intelligence. The method includes: inputting data of a strategy game into a pre-trained artificial intelligence model corresponding to the current student to output a first competitive strategy, generating a second competitive strategy through the input instructions of the current student for the strategy game, and then respectively controlling the two sides' characters to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current student; inputting data of the strategy game into a pre-trained artificial intelligence model corresponding to other students to output a third competitive strategy, and then respectively controlling the two sides' characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current student; adjusting the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guiding the current student to enter course learning.

[0227] In addition, when the logic instructions in the above-mentioned memory 830 are implemented in the form of software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, may be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0228] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the artificial intelligence-based interactive teaching method provided by each of the above methods. The method includes: inputting the data of a strategy game into a pre-trained artificial intelligence model corresponding to the current student to output a first competitive strategy, generating a second competitive strategy through the input instructions of the current student for the strategy game, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current student; inputting the data of the strategy game into a pre-trained artificial intelligence model corresponding to other students to output a third competitive strategy, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current student; adjusting the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guiding the current student to enter the course study.

[0229] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the artificial intelligence-based interactive teaching method provided by each of the above methods. The method includes: inputting the data of a strategy game into a pre-trained artificial intelligence model corresponding to the current student to output a first competitive strategy, generating a second competitive strategy through the input instructions of the current student for the strategy game, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the second competitive strategy to obtain the first score of the current student; inputting the data of the strategy game into a pre-trained artificial intelligence model corresponding to other students to output a third competitive strategy, and then respectively controlling the two characters to conduct a strategy game battle based on the first competitive strategy and the third competitive strategy to obtain the second score of the current student; adjusting the curriculum syllabus and learning plan according to the first score, the second score, and the data of the strategy game, and guiding the current student to enter the course study.

[0230] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0231] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0232] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An interactive teaching method based on artificial intelligence, characterized in that: For use in a teaching system, the method comprises: By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, a first competitive strategy is output, a second competitive strategy is generated through the input instructions of the current student for the strategy game, and then based on the first competitive strategy and the second competitive strategy, the characters of both parties are controlled to play the strategy game, and a first score of the current student is obtained; By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to other students, outputting a third competitive strategy, and then controlling the characters of both parties to play the strategy game based on the first competitive strategy and the third competitive strategy, a second score of the current student is obtained; Adjusting the course outline and study plan according to the first score, the second score and the data of the strategy game, and guiding the current students to start course study; The AI ​​model corresponding to the current students is trained in the following ways: Training data screening: Real-time capture and recording of the strategies and results adopted by students after the game, and storing these data in a secure data storage module, and automatically selecting the most recent N wins as the winning strategy data; Training data enhancement: Using data enhancement algorithms, more sample data is automatically generated from winning strategy data, including winning sample data and losing sample data; Feature engineering: Calculate the difference between the coordinate values ​​of game props to generate new data points to expand the feature dimension of the data set; Algorithm selection: According to the conditions of the students, select the artificial intelligence algorithm that matches the students; Parameter adjustment: adjust the algorithm parameters according to the type of selected algorithm; Model building and optimization: Use the screened training data, selected algorithms and parameters to train the artificial intelligence model and save it to the operation and maintenance module.

2. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: The data of the strategy game includes the number of props and the position of props; By inputting the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, the first competitive strategy is output, including: According to the number and location of props in the strategy game, multiple competitive strategies for obtaining props are generated; wherein each competitive strategy includes the competition order and path of multiple props; The multiple competitive strategies are input into the artificial intelligence model corresponding to the current student for calculation to obtain the winning probability corresponding to each competitive strategy, and the competitive strategy with the highest winning probability is output as the first competitive strategy.

3. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: Generate the second competitive strategy through the current student's input instructions, including: Determine the competition order and path of the props displayed on the screen through the current student's input instructions; A second competition strategy is generated according to the competition order and path of the props.

4. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: The data of the strategy game includes the number of props and the position of props; By inputting the data of the strategy game into the corresponding pre-trained artificial intelligence models of other students, the third competitive strategy is output, including: According to the number and location of props in the strategy game, multiple competitive strategies for obtaining props are generated; wherein each competitive strategy includes the competition order and path of multiple props; The multiple competitive strategies are input into the artificial intelligence models corresponding to other students for calculation, so as to obtain the winning probability corresponding to each competitive strategy, and the competitive strategy with the highest winning probability is output as the third competitive strategy.

5. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: Based on the first competitive strategy and the second competitive strategy, the characters of both parties are controlled to play a strategic game against each other, and a first score of the current student is obtained, including: Based on the first competition strategy and the second competition strategy, the characters of both parties are controlled to obtain props according to the competition order and path of their respective props; If one of the characters obtains any item first, the item will be removed from the competition order and path of both characters, and their respective competition order and path will be updated; Continue to control the two characters to continue to obtain props according to the updated order and path of the competition for props, until all props are obtained, and count the number of props obtained by each character; Determine the current student's first score based on the difference in the number of props.

6. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: Based on the first competitive strategy and the third competitive strategy, the characters of both parties are controlled to play a strategic game battle, and a second score of the current student is obtained, specifically including: Based on the first competition strategy and the third competition strategy, the characters of both parties are controlled to obtain props according to the competition order and path of their respective props; If one of the characters obtains any item first, the item will be removed from the competition order and path of both characters, and their respective competition order and path will be updated; Continue to control the two characters to continue to obtain props according to the updated order and path of the competition for props, until all props are obtained, and count the number of props obtained by each character; The second score of the current student is determined by calculating based on the initial score of the current student's character, the initial scores of other students' characters, the number of props obtained by the current student, and the number of props obtained by other students.

7. The interactive teaching method based on artificial intelligence according to claim 1, characterized in that: Adjusting the course outline and study plan according to the first score, the second score, and the data of the strategy game, and guiding the current students to enter the course study, including: If the score obtained by the current student in the first competitive strategy exceeds the first set score threshold, a competitive mode is unlocked in which the first competitive strategy and the second competitive strategy are used to control the two characters to play a strategy game, and basic artificial intelligence popular science teaching content is added to the current student's course outline; In the case where the first score of the current student exceeds the second set score threshold, adding an explanation of the basic principles of artificial intelligence to the course outline of the current student; In the case where the second score of the current student exceeds the third set score threshold, continue to add relevant knowledge of establishing an artificial intelligence model to the course outline of the current student.

8. An interactive teaching device based on artificial intelligence, characterized in that: For use in a teaching system, the device comprises: The human-machine battle module is used to input the data of the strategy game into the pre-trained artificial intelligence model corresponding to the current student, output a first competitive strategy, generate a second competitive strategy through the input instructions of the current student for the strategy game, and then control the characters of both parties to play the strategy game based on the first competitive strategy and the second competitive strategy, and obtain a first score of the current student; The machine battle module is used to input the data of the strategy game into the pre-trained artificial intelligence model corresponding to other students, output a third competitive strategy, and then control the characters of both sides to play the strategy game based on the first competitive strategy and the third competitive strategy, so as to obtain the second score of the current student; A course selection module, used for adjusting the course outline and study plan according to the first score, the second score and the data of the strategy game, and guiding the current student to enter the course study; The AI ​​model corresponding to the current students is trained in the following ways: Training data screening: Real-time capture and recording of the strategies and results adopted by students after the game, and storing these data in a secure data storage module, and automatically selecting the most recent N wins as the winning strategy data; Training data enhancement: Using data enhancement algorithms, more sample data is automatically generated from winning strategy data, including winning sample data and losing sample data; Feature engineering: Calculate the difference between the coordinate values ​​of game props to generate new data points to expand the feature dimension of the data set; Algorithm selection: According to the conditions of the students, select the artificial intelligence algorithm that matches the students; Parameter adjustment: adjust the algorithm parameters according to the type of selected algorithm; Model building and optimization: Use the screened training data, selected algorithms and parameters to train the artificial intelligence model and save it to the operation and maintenance module.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the interactive teaching method based on artificial intelligence as described in any one of claims 1 to 7 is implemented.

Images