Foreign language situational teaching intelligent management system and method based on virtual reality

Abstract

The invention relates to the technical field of virtual reality, in particular to a foreign language situational teaching intelligent management system and method based on virtual reality, in the system, a dynamic learning portrait generation module comprises a semantic feature extraction unit, a language expression feature extraction unit and a portrait construction unit, user dialogue information is obtained in real time, and semantic features of a user are extracted based on a language recognition technology; the language expression feature extraction unit obtains user dialogue information in real time in the situational teaching process and extracts language expression features of a user based on the eye movement tracking technology. According to the semantic feature and language expression feature extraction result of the user dialogue information, the dynamic learning portrait of the user is constructed, the differentiation demand of the user is analyzed, and the scene group learning thermodynamic distribution diagram in the historical data is combined, so that the dialogue content in the foreign language teaching process is automatically adjusted.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual reality, and in particular to an intelligent management system and method for foreign language scenario teaching based on virtual reality. Background Art

[0002] With the acceleration of the globalization process, the importance of foreign language application ability has become increasingly prominent. Traditional foreign language teaching mostly relies on classroom lectures and static teaching materials, lacking interaction and practice in real contexts, resulting in students' difficulty in flexibly applying language skills. Although virtual reality (VR) technology provides a new way for immersive language learning, existing systems still have significant limitations and urgently need to achieve breakthroughs through intelligent management.

[0003] Traditional teaching is limited by the physical environment and is difficult to simulate dynamic real scenarios; while early VR foreign language teaching systems improved the immersion by presetting scenarios (such as ordering food in a restaurant), but had the following deficiencies, for example: the scenario library is single and static, unable to dynamically generate diverse scenarios; and not adapting to adjust the difficulty in combination with the learner's level, progress, and cognitive characteristics, making it difficult to meet the differentiated needs of users, etc.; therefore, the existing intelligent management systems for foreign language scenario teaching based on virtual reality have relatively large defects. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent management system and method for foreign language scenario teaching based on virtual reality to solve the problems raised in the above background art.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: An intelligent management method for foreign language scenario teaching based on virtual reality, including:

[0006] Step S100: Extract keywords in the human-computer interaction information in the scene construction stage, match the dialogue scenarios and virtual characters in the database, and generate a virtual scenario model;

[0007] Step S200: Based on the heat distribution map of scene group learning in historical data, retrieve an alternative set of scenario teaching information based on the virtual scenario model;

[0008] Step S300: During the scenario teaching process, obtain the user's dialogue information in real time, extract the semantic features of the user based on speech recognition technology, extract the language expression features of the user based on eye movement tracking technology, and generate a dynamic learning portrait of the user;

[0009] Step S400: Generate the teaching preference of the user based on the virtual scenario model based on the dynamic learning portrait of the user; combine the teaching preference of the user based on the virtual scenario model to dynamically update the alternative set of scenario teaching information based on the virtual scenario model.

[0010] Further, the step S100 includes:

[0011] Obtain the human-computer interaction information during the scenario construction stage. Based on the preset keywords in the preset form in the database, extract keywords from the human-computer interaction information during the scenario construction stage. Each preset keyword or combination of multiple preset keywords extracted corresponds to one or more dialogue scenario elements and corresponding virtual characters in the database;

[0012] The virtual scenario model is composed of a matching dialogue scenario and virtual characters. The matching dialogue scenario in the virtual scenario model is a set of dialogue scenario elements corresponding to the keyword extraction result of the human-computer interaction information during the scenario construction stage; the matching virtual characters in the virtual scenario model are a set of each virtual character corresponding to the keyword extraction result of the human-computer interaction information during the scenario construction stage.

[0013] The present invention can identify the dialogue scenario that the user needs to simulate according to the human-computer interaction data, and identify the dynamic matching of the dialogue scenario elements according to the difference of the identified keywords, so as to ensure a high degree of adaptation between the constructed virtual scenario model and the user's dialogue content.

[0014] Further, the step S200 includes:

[0015] Step S201, obtain the scenario group learning heat distribution map under each virtual scenario model in the historical data; the scenario group learning heat distribution map includes the knowledge points with abnormal user interactions in the scenario teaching data under the corresponding virtual scenario model and the ratio of the abnormal frequency of each knowledge point to the total abnormal frequency of all knowledge points;

[0016] Step S202, obtain the generated virtual scenario model, and extract the scenario group learning heat distribution map under the virtual scenario model with the highest similarity to the generated virtual scenario model in the historical data, denoted as the comparison scenario group learning heat distribution map; the formula for calculating the similarity between different virtual scenario models is as follows:

[0017] SM i =NDQ i / NDW i +σ·NXQ i / max{NXW i ,NXW}

[0018] Among them, SM i represents the similarity between the i-th virtual scenario model in the historical data and the generated virtual scenario model; NDQ i represents the number of dialogue scenario elements in the intersection of the dialogue scenario matched by the i-th virtual scenario model in the historical data and the dialogue scenario matched by the generated virtual scenario model; NDW irepresents the number of dialogue scene elements in the union of the dialogue scenes matched by the i-th virtual scenario model in the historical data and the dialogue scenes matched by the generated virtual scenario model; NXQ i represents the number of identical virtual characters in the i-th virtual scenario model in the historical data and the generated virtual scenario model; NXW i represents the number of virtual characters in the i-th virtual scenario model in the historical data; NXW represents the number of virtual characters in the generated virtual scenario model; σ represents a preset weight coefficient; max{} represents the operation of finding the maximum value;

[0019] Step S203: The summary set of the top n knowledge points in the ranking of knowledge points in descending order of the ratio of the corresponding abnormal frequency in the contrast scene group learning heat distribution map to the total abnormal frequency of all knowledge points is denoted as the alternative set of scenario teaching information based on the virtual scenario model; the n is a preset value; and the summary set of communication topics involved in each knowledge point in the contrast scene group learning heat distribution map in the historical data is bound to the corresponding knowledge point; the communication topics bound to the knowledge points are obtained by matching the communication topics corresponding to each preset keyword in the database preset form in the scenario dialogue content.

[0020] In the present invention, the knowledge points involved in the scenario teaching data corresponding to different users under the same virtual scenario model are different, and the types of knowledge points involved in the scenario group learning heat distribution map under the virtual scenario model are the summary set of the types of interaction abnormal knowledge points involved in the scenario teaching data corresponding to different users under the corresponding virtual scenario model.

[0021] Further, the step S300 includes:

[0022] The semantic features of the user include the semantic recognition results of each word and sentence in the user's dialogue information, and the words or sentences with abnormal semantic recognition results;

[0023] The facial expressions of the user during the dialogue are captured in real time through a camera, and the eye information of the user is tracked to generate a language expression feature based on the user's dialogue information. The language expression feature includes the summary set of the corresponding facial expression types of the facial expression capture results corresponding to each dialogue sentence of the user, the anxiety coefficient corresponding to the expression process of each dialogue sentence, and the eye movement frequency during the expression process of each dialogue sentence.

[0024] The facial expression type corresponding to the facial expression capture result is obtained by performing face recognition on the facial expression capture result; each facial expression type in the database corresponds to a preset anxiety coefficient, and the anxiety coefficient corresponding to each dialogue statement expression process is equal to the average value of the anxiety coefficients corresponding to the facial expression types to which each facial expression captured in the facial expression capture result corresponding to the dialogue statement belongs; the eye movement frequency during each dialogue statement expression process is the ratio of the number of user eyeball movement cycles to the total duration of the corresponding dialogue statement expression, and the eyeball movement directions at each time point during the user's eyeball movement cycle are the same and the maximum movement distance of the eyeball is greater than the preset distance;

[0025] Perform weighted calculation on the anxiety coefficient and the corresponding eye movement frequency corresponding to the user during the expression process of the same dialogue statement in the language expression characteristics, and the weighting coefficient in the weighted calculation process is a preset value, and take the obtained weighted calculation result as the fluency corresponding to the user's dialogue statement expression process; screen the extraction times of each facial expression corresponding to the anxiety coefficient greater than the anxiety coefficient corresponding to the dialogue statement expression process in the dialogue statements with fluency less than or equal to the preset fluency, and perform secondary semantic anomaly recognition on the words expressed by the obtained extraction times during the dialogue statement expression process with fluency less than or equal to the preset fluency, and update the user's semantic characteristics according to the secondary semantic anomaly recognition result;

[0026] The dynamic learning portrait of the user is composed of the user's language expression characteristics and the updated user semantic characteristics.

[0027] The present invention monitors the user's scenario dialogue state from multiple angles of speech recognition, facial capture emotion recognition and eye movement tracking, which is convenient for analyzing the adaptation situation between the scenario dialogue content and the user, provides a data basis for dynamically analyzing the teaching preference of the user based on the virtual scenario model subsequently, and dynamically updating the scenario teaching information alternative set based on the virtual scenario model.

[0028] Further, the method for generating the teaching preference of the user based on the virtual scenario model in step S400 includes:

[0029] Step S401, obtain the dynamic learning portrait of the user;

[0030] Step S402, obtain each communication topic in the user dialogue information and the topic content corresponding to each communication topic; regard each dialogue statement between two adjacent different communication topics in the user dialogue information as the topic content of the previous communication topic in the two;

[0031] Step S403, calculate the deviation adaptation value of the user's dynamic learning portrait based on each communication topic in the user dialogue information, and the involved calculation formula is as follows:

[0032]

[0033] Among them, P m represents the deviation adaptation value of the user's dynamic learning portrait based on the m-th communication topic in the user's conversation information; BY m represents the ratio of the number of abnormal elements in the semantic recognition results within the semantic features belonging to the m-th communication topic content in the user's dynamic learning portrait to the total number of abnormal elements in the semantic recognition results within the semantic feature ratio in the user's dynamic learning portrait; BL (m，k) represents the anxiety coefficient corresponding to the expression process of the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning portrait; Km represents the number of dialogue statements belonging to the m-th communication topic content in the user's dynamic learning portrait; BD (m，k) represents the eye movement frequency during the expression process of the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning portrait; BC (m，k) represents the vocabulary in the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning portrait; μ1 represents a preset normalization coefficient; μ2 represents a preset second normalization coefficient;

[0034] Step S404: Use the summary set of all communication topics whose deviation coefficient between the corresponding deviation adaptation value and the maximum deviation adaptation value of the user's dynamic learning portrait based on the communication topics in the user's conversation information is less than the preset deviation value as the teaching preference of the user based on the virtual scenario model; the deviation coefficient between the corresponding deviation adaptation value and the maximum deviation adaptation value of the user's dynamic learning portrait based on the communication topics in the user's conversation information is equal to the quotient obtained by dividing the difference between the maximum deviation adaptation value of the user's dynamic learning portrait based on the communication topics in the user's conversation information and the corresponding deviation adaptation value by the maximum deviation adaptation value of the user's dynamic learning portrait based on the communication topics in the user's conversation information.

[0035] Furthermore, during the process of dynamically updating the alternative set of scenario teaching information based on the virtual scenario model in step S400,

[0036] Obtain the teaching preference of the user based on the virtual scenario model, eliminate each knowledge point in the alternative set of scenario teaching information based on the virtual scenario model whose intersection with the summary set of communication topics corresponding to the binding is empty, and obtain the updated result of the alternative set of scenario teaching information based on the virtual scenario model. Moreover, each element in the summary set of communication topics bound to each element in the updated alternative set of scenario teaching information based on the virtual scenario model belongs to the teaching preference of the user based on the virtual scenario model.

[0037] A foreign language scenario teaching intelligent management system based on virtual reality, which includes a virtual scenario model building module, a scenario group learning heat distribution analysis module, a dynamic learning portrait generation module and a scenario teaching content management module;

[0038] The virtual scenario model building module extracts keywords from the human-computer interaction information in the scenario construction stage, matches the dialogue scenes and virtual characters in the database, and generates a virtual scenario model; the scenario group learning heat distribution analysis module retrieves the scenario teaching information alternative set based on the virtual scenario model based on the scenario group learning heat distribution map in the historical data; the dynamic learning portrait generation module obtains user dialogue information in real time during the scenario teaching process, and extracts the user's semantic features based on language recognition technology, and extracts the user's language expression features based on eye tracking technology to generate the user's dynamic learning portrait; the scenario teaching content management module generates the user's teaching preference based on the virtual scenario model based on the user's dynamic learning portrait; and dynamically updates the scenario teaching information alternative set based on the virtual scenario model in combination with the user's teaching preference based on the virtual scenario model.

[0039] Furthermore, the dynamic learning portrait generation module includes a semantic feature extraction unit, a language expression feature extraction unit and a portrait construction unit.

[0040] The semantic feature extraction unit obtains user conversation information in real time during the situational teaching process and extracts the user's semantic features based on language recognition technology;

[0041] The language expression feature extraction unit obtains user dialogue information in real time during the situational teaching process and extracts the user's language expression features based on eye tracking technology;

[0042] The portrait construction unit generates a dynamic learning portrait of the user based on the acquisition results of the semantic feature extraction unit and the language expression feature extraction unit.

[0043] Furthermore, the scenario teaching content management module includes a bias analysis unit and a data update unit.

[0044] The bias analysis unit generates the user's teaching bias based on the virtual scenario model based on the user's dynamic learning profile;

[0045] The data updating unit dynamically updates the scenario teaching information selection set based on the virtual scenario model in combination with the user's teaching preference based on the virtual scenario model.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) The present invention realizes the dynamic construction of a foreign language scenario teaching scenario by extracting keywords in the human-computer interaction information during the scenario construction stage.

[0048] (2) The present invention constructs a dynamic learning portrait of the user according to the semantic features and language expression feature extraction results of the user's dialogue information, analyzes the user's differentiated needs, and combines the scenario group learning heat distribution map in the historical data to realize the automatic adjustment of the dialogue content during the foreign language teaching process and effectively manage the foreign language scenario teaching data. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0050] Figure 1 is a schematic flow chart of the intelligent management system for foreign language scenario teaching based on virtual reality according to the present invention;

[0051] Figure 2 is a schematic structural diagram of the intelligent management method for foreign language scenario teaching based on virtual reality according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0053] The present invention provides a technical solution: as Figure 1 shown, the intelligent management system for foreign language scenario teaching based on virtual reality in this embodiment includes a virtual scenario model building module, a scenario group learning heat distribution analysis module, a dynamic learning portrait generation module, and a scenario teaching content management module;

[0054] The virtual scenario model building module extracts keywords in the human-computer interaction information during the scenario construction stage, matches the dialogue scenarios and virtual characters in the database, and generates a virtual scenario model;

[0055] The scenario group learning heat distribution analysis module retrieves an alternative set of scenario teaching information based on the virtual scenario model based on the scenario group learning heat distribution map in the historical data;

[0056] The dynamic learning portrait generation module includes a semantic feature extraction unit, a language expression feature extraction unit, and a portrait construction unit,

[0057] During the scenario-based teaching process, the semantic feature extraction unit obtains the user's dialogue information in real time and extracts the semantic features of the user based on language recognition technology;

[0058] During the scenario-based teaching process, the language expression feature extraction unit obtains the user's dialogue information in real time and extracts the language expression features of the user based on eye movement tracking technology;

[0059] The portrait construction unit generates a dynamic learning portrait of the user according to the acquisition results of the semantic feature extraction unit and the language expression feature extraction unit.

[0060] The scenario-based teaching content management module includes a bias analysis unit and a data update unit.

[0061] The bias analysis unit generates a teaching bias of the user based on the virtual scenario model based on the user's dynamic learning portrait.

[0062] The data update unit dynamically updates the alternative set of scenario-based teaching information based on the virtual scenario model in combination with the teaching bias of the user based on the virtual scenario model.

[0063] As Figure 2 shown, the intelligent management method for virtual reality-based foreign language scenario teaching in this embodiment includes:

[0064] Step S100: Extract keywords in the human-computer interaction information in the scene construction stage, match the dialogue scenarios and virtual characters in the database, and generate a virtual scenario model;

[0065] In the step S100, it includes:

[0066] Obtain the human-computer interaction information in the scene construction stage, extract keywords from the human-computer interaction information in the scene construction stage based on the preset keywords in the preset form in the database, and each preset keyword or a combination of multiple preset keywords in the database corresponds to one or more dialogue scenario elements and corresponding virtual characters respectively;

[0067] The virtual scenario model is composed of the matched dialogue scenarios and virtual characters. The matched dialogue scenarios in the virtual scenario model are a set of dialogue scenario elements corresponding to the keyword extraction results of the human-computer interaction information in the scene construction stage; the matched virtual characters in the virtual scenario model are a set of each virtual character corresponding to the keyword extraction results of the human-computer interaction information in the scene construction stage.

[0068] Step S200: Retrieve the alternative set of scenario-based teaching information based on the virtual scenario model based on the scene group learning heat distribution map in the historical data;

[0069] In the step S200, it includes:

[0070] Step S201: Obtain the scene group learning heat distribution maps under each virtual scenario model in the historical data; the scene group learning heat distribution maps include the knowledge points with abnormal user interactions in the scenario teaching data under the corresponding virtual scenario model and the ratio of the abnormal frequency of each knowledge point to the total abnormal frequency of all knowledge points; the knowledge points involved in the scenario teaching data corresponding to different users under the same virtual scenario model are different, and the types of knowledge points involved in the scene group learning heat distribution map under the virtual scenario model are the aggregated set of the types of interaction abnormal knowledge points involved in the scenario teaching data corresponding to different users under the corresponding virtual scenario model;

[0071] Step S202: Obtain the generated virtual scenario model, and extract the scene group learning heat distribution map under the virtual scenario model with the highest similarity to the generated virtual scenario model in the historical data, denoted as the comparison scene group learning heat distribution map; the formula for calculating the similarity between different virtual scenario models is as follows:

[0072] SM i =NDQ i / DNW i +σ·NXQ i / max{NXW i ,NXW}

[0073] where, SM i represents the similarity between the i-th virtual scenario model in the historical data and the generated virtual scenario model; NDQ i represents the number of dialogue scene elements in the intersection of the dialogue scenes matched by the i-th virtual scenario model in the historical data and the dialogue scenes matched by the generated virtual scenario model; NDW i represents the number of dialogue scene elements in the union of the dialogue scenes matched by the i-th virtual scenario model in the historical data and the dialogue scenes matched by the generated virtual scenario model; NXQ i represents the number of identical virtual characters in the i-th virtual scenario model and the generated virtual scenario model in the historical data; NXW i represents the number of virtual characters in the i-th virtual scenario model in the historical data; NXW represents the number of virtual characters in the generated virtual scenario model; σ represents a preset weight coefficient; max{} represents the operation of finding the maximum value;

[0074] Step S203: Sort the ratios of the abnormal frequencies corresponding to the knowledge points in the comparison scenario group learning heat distribution map to the total abnormal frequencies of all knowledge points from largest to smallest, and summarize the top n knowledge points in the sorted order. Denote this as the alternative set of scenario teaching information based on the virtual scenario model; where n is a preset value; and bind the summary set of communication topics involved in each knowledge point in the comparison scenario group learning heat distribution map in the historical data to the corresponding knowledge points; the communication topics bound to the knowledge points are obtained by matching the corresponding communication topics of each preset keyword in the scenario dialogue content in the database preset form.

[0075] Step S300: During the scenario teaching process, obtain the user's dialogue information in real time, extract the semantic features of the user based on speech recognition technology, extract the language expression features of the user based on eye movement tracking technology, and generate a dynamic learning portrait of the user.

[0076] The step S300 includes:

[0077] The semantic features of the user include the semantic recognition results of each word and sentence in the user's dialogue information, and the words or sentences with abnormal semantic recognition results.

[0078] Capture the user's facial expressions in real time during the dialogue through a camera, and track the user's eye information to generate language expression features based on the user's dialogue information. The language expression features include the summary set of facial expression types corresponding to the facial expression capture results for each dialogue sentence, the anxiety coefficient corresponding to each dialogue sentence expression process, and the eye movement frequency during each dialogue sentence expression process.

[0079] The facial expression types corresponding to the facial expression capture results are obtained through face recognition of the facial expression capture results; each facial expression type in the database corresponds to a preset anxiety coefficient, and the anxiety coefficient corresponding to each dialogue sentence expression process is equal to the average value of the anxiety coefficients corresponding to the facial expression types to which each facial expression captured in the facial expression capture results corresponding to the dialogue sentence belongs; the eye movement frequency during each dialogue sentence expression process is the ratio of the number of eye movement cycles of the user during the dialogue sentence expression process to the total duration of the dialogue sentence expression, and the eye movement directions at each time point in the user's eye movement cycle are the same and the maximum movement distance of the eyeball is greater than the preset distance.

[0080] The foreign language taught in this embodiment includes English.

[0081] Perform weighted calculation on the anxiety coefficient and corresponding eye movement frequency corresponding to the user during the expression process of the same dialogue statement in the language expression characteristics, and the weighting coefficient in the weighted calculation process is a preset value, and use the obtained weighted calculation result as the fluency corresponding to the user's corresponding dialogue statement expression process; screen the extraction times of each facial expression corresponding to the anxiety coefficient greater than the anxiety coefficient corresponding to the corresponding dialogue statement expression process in the dialogue statements with fluency less than or equal to the preset fluency, and perform secondary semantic anomaly recognition on the words expressed by the obtained extraction times respectively during the expression process of the dialogue statements with fluency less than or equal to the preset fluency, and update the semantic characteristics of the user according to the secondary semantic anomaly recognition result;

[0082] The dynamic learning portrait of the user is composed of the user's language expression characteristics and the updated user semantic characteristics.

[0083] Step S400: Generate the teaching preference of the user based on the virtual scenario model according to the dynamic learning portrait of the user; combine the teaching preference of the user based on the virtual scenario model to dynamically update the alternative set of scenario teaching information based on the virtual scenario model;

[0084] The method for generating the teaching preference of the user based on the virtual scenario model in the step S400 includes:

[0085] Step S401: Obtain the dynamic learning portrait of the user;

[0086] Step S402: Obtain each communication topic in the user's dialogue information and the topic content corresponding to each communication topic; regard each dialogue statement between two adjacent different communication topics in the user's dialogue information as the topic content of the previous communication topic among the two;

[0087] Step S403: Calculate the bias adaptation value of the user's dynamic learning portrait based on each communication topic in the user's dialogue information, and the involved calculation formula is as follows:

[0088]

[0089] where, P m represents the bias adaptation value of the user's dynamic learning portrait based on the mth communication topic in the user's dialogue information; BY m represents the ratio of the number of semantic recognition result abnormal elements in the semantic features belonging to the mth communication topic content in the user's dynamic learning portrait to the total number of semantic recognition result abnormal elements in the semantic feature ratio in the user's dynamic learning portrait; BL (m，k) represents the anxiety coefficient corresponding to the expression process of the kth dialogue statement belonging to the mth communication topic content in the user's dynamic learning portrait; Km represents the number of dialogue statements belonging to the mth communication topic content in the user's dynamic learning portrait; BD (m，k)represents the eye movement frequency during the expression of the kth dialogue sentence belonging to the mth communication topic in the user's dynamic learning profile; BC (m，k) represents the vocabulary size in the kth conversation sentence belonging to the mth communication topic in the user's dynamic learning profile; μ1 represents the preset normalization coefficient; μ2 represents the preset second normalization coefficient;

[0090] Step S404: A summary set of all communication topics for which the deviation coefficient between the corresponding bias adaptation value and the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information is less than a preset deviation value is used as the user's teaching bias based on the virtual scenario model; the deviation coefficient between the corresponding bias adaptation value and the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information is equal to the quotient of the difference between the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information and the corresponding bias adaptation value divided by the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information.

[0091] In the process of dynamically updating the scenario teaching information alternative set based on the virtual scenario model in step S400, the user's teaching bias based on the virtual scenario model is obtained, and the knowledge points whose intersection between the corresponding bound communication topic summary set in the scenario teaching information alternative set based on the virtual scenario model and the user's teaching bias based on the virtual scenario model are empty are eliminated to obtain the update result of the scenario teaching information alternative set based on the virtual scenario model, and the elements of the communication topic summary set bound to each element in the updated scenario teaching information alternative set based on the virtual scenario model all belong to the user's teaching bias based on the virtual scenario model.

[0092] In this embodiment, the elements in the communication topic summary set bound to each element in the updated scenario teaching information alternative set based on the virtual scenario model that do not belong to the user's teaching preference based on the virtual scenario model are eliminated, so that the elements in the communication topic summary set bound to each element in the updated scenario teaching information alternative set based on the virtual scenario model all belong to the user's teaching preference based on the virtual scenario model.

[0093] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0094] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements on some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent management method for foreign language situational teaching based on virtual reality, characterized in that, Including: Step S100: Extract keywords from the human-computer interaction information in the scenario construction stage, match the dialogue scenarios and virtual characters in the database, and generate a virtual scenario model; Step S200: Based on the heat distribution map of scenario group learning in historical data, retrieve the alternative set of scenario teaching information based on the virtual scenario model; Step S300: During the scenario teaching process, obtain the user's dialogue information in real time, extract the semantic features of the user based on speech recognition technology, extract the language expression features of the user based on eye movement tracking technology, and generate a dynamic learning portrait of the user; Step S400: Generate the teaching preference of the user based on the virtual scenario model based on the dynamic learning portrait of the user; Combined with the teaching preference of the user based on the virtual scenario model, dynamically update the alternative set of scenario teaching information based on the virtual scenario model.

2. The intelligent management method for foreign language situational teaching based on virtual reality according to claim 1, characterized in that: In the step S100, it includes: Obtain the human-computer interaction information in the scenario construction stage, and extract keywords from the human-computer interaction information in the scenario construction stage based on the preset keywords in the preset form in the database. Each preset keyword or combination of multiple preset keywords extracted corresponds to one or more dialogue scenario elements and corresponding virtual characters in the database; The virtual scenario model is composed of the matched dialogue scenario and virtual characters. The matched dialogue scenario in the virtual scenario model is a set of dialogue scenario elements corresponding to the keyword extraction result of the human-computer interaction information in the scenario construction stage; the matched virtual characters in the virtual scenario model are a set of each virtual character corresponding to the keyword extraction result of the human-computer interaction information in the scenario construction stage.

3. The intelligent management method for foreign language situational teaching based on virtual reality according to claim No. 1, wherein: In the step S200, it includes: Step S201: Obtain the heat distribution map of scenario group learning under each virtual scenario model in historical data; the heat distribution map of scenario group learning includes the knowledge points with abnormal user interaction in the scenario teaching data under the corresponding virtual scenario model and the ratio of the abnormal frequency of each knowledge point to the total abnormal frequency of all knowledge points; Step S202: Obtain the generated virtual scenario model, and extract the heat distribution map of scenario group learning under the virtual scenario model with the highest similarity to the generated virtual scenario model in historical data, denoted as the comparison heat distribution map of scenario group learning; the formula for calculating the similarity between different virtual scenario models is as follows: SM i = NDQ i / NDW i + σ·NXQ i / max{NXW i , NXW} Among them, SM i represents the similarity between the i-th virtual scenario model in the historical data and the generated virtual scenario model; NDQ i represents the number of dialogue scenario elements in the intersection of the dialogue scenarios matched by the i-th virtual scenario model in the historical data and the dialogue scenarios matched by the generated virtual scenario model; NDW i represents the number of dialogue scenario elements in the union of the dialogue scenarios matched by the i-th virtual scenario model in the historical data and the dialogue scenarios matched by the generated virtual scenario model; NXQ i represents the number of identical virtual characters in the i-th virtual scenario model in the historical data and the generated virtual scenario model; NXW i represents the number of virtual characters in the i-th virtual scenario model in the historical data; NXW represents the number of virtual characters in the generated virtual scenario model; σ represents a preset weight coefficient; max{} represents the operation of finding the maximum value; Step S203: Denote the summary set of the first n knowledge points in the ranking of knowledge points by the ratio of the corresponding abnormal frequency to the total abnormal frequency of all knowledge points in the comparison heat distribution map of scenario group learning as the alternative set of scenario teaching information based on the virtual scenario model; the n is a preset value; and bind the summary set of communication topics involved in each knowledge point in the comparison heat distribution map of scenario group learning in historical data to the corresponding knowledge points; the communication topics bound to the knowledge points are obtained by matching the communication topics corresponding to each preset keyword in the preset form in the database in the scenario dialogue content.

4. The intelligent management method for foreign language situational teaching based on virtual reality according to claim 1, characterized in that: The step S300 includes: The semantic features of the user include the semantic recognition results of each word and sentence in the user's dialogue information, and the words or sentences with abnormal semantic recognition results; The facial expressions of the user during the conversation are captured in real time through a camera, and the eye information of the user is tracked to generate language expression features based on the user's conversation information. The language expression features include a summary set of facial expression types corresponding to the facial expression capture results corresponding to each conversation statement of the user, the anxiety coefficient corresponding to each conversation statement expression process, and the eye movement frequency during each conversation statement expression process. The facial expression types corresponding to the facial expression capture results are obtained by performing face recognition on the facial expression capture results; each facial expression type in the database corresponds to a preset anxiety coefficient, and the anxiety coefficient corresponding to each conversation statement expression process is equal to the average value of the anxiety coefficients corresponding to the facial expression types to which each facial expression captured in the facial expression capture results corresponding to the corresponding conversation statement belongs; the eye movement frequency during each conversation statement expression process is the ratio of the number of eyeball movement cycles of the user during the corresponding conversation statement expression process to the total duration of the corresponding conversation statement expression, and the eyeball movement directions at each time point during the user's eyeball movement cycle are the same and the maximum movement distance of the eyeball is greater than the preset distance. Perform weighted calculation on the anxiety coefficient and the corresponding eye movement frequency corresponding to the user during the expression process of the same conversation statement in the language expression features, and the weighting coefficient in the weighted calculation process is a preset value. Take the obtained weighted calculation result as the fluency corresponding to the user during the expression process of the corresponding conversation statement; screen the extraction times of each facial expression corresponding to the anxiety coefficient greater than the anxiety coefficient corresponding to the corresponding conversation statement expression process in the conversation statements with fluency less than or equal to the preset fluency, and perform secondary semantic anomaly recognition on the words expressed by the obtained extraction times respectively during the expression process of the conversation statements with fluency less than or equal to the preset fluency, and update the semantic features of the user according to the secondary semantic anomaly recognition results. The dynamic learning portrait of the user is composed of the user's language expression features and the updated user semantic features.

5. The intelligent management method for foreign language scenario teaching based on virtual reality according to claim 4, characterized in that: The method for generating the teaching preference of the user based on the virtual scenario model in step S400 includes: Step S401, obtain the dynamic learning portrait of the user; Step S402, obtain each communication topic in the user's conversation information and the topic content corresponding to each communication topic; regard each conversation statement between two adjacent different communication topics in the user's conversation information as the topic content of the previous communication topic of the two. Step S403, calculate the preference adaptation value of the user's dynamic learning portrait based on each communication topic in the user's conversation information. The involved calculation formula is as follows: Among them, P m represents the deviation adaptation value of the user's dynamic learning profile based on the m-th communication topic in the user conversation information; BY m represents the ratio of the number of abnormal semantic recognition elements in the semantic features belonging to the m-th communication topic content in the user's dynamic learning profile to the total number of abnormal semantic recognition elements in the semantic feature ratio in the user's dynamic learning profile; BL (m，k) represents the anxiety coefficient corresponding to the expression process of the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning profile; Km represents the number of dialogue statements belonging to the m-th communication topic content in the user's dynamic learning profile; BD (m，k) represents the eye movement frequency during the expression process of the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning profile; BC (m，k) represents the vocabulary in the k-th dialogue statement belonging to the m-th communication topic content in the user's dynamic learning profile; μ1 represents a preset normalization coefficient; μ2 represents a preset second normalization coefficient; Step S404: A summary set of all communication topics for which the deviation coefficient between the corresponding bias adaptation value and the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information is less than a preset deviation value is used as the user's teaching bias based on the virtual scenario model; the deviation coefficient between the corresponding bias adaptation value and the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information is equal to the quotient of the difference between the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information and the corresponding bias adaptation value divided by the maximum bias adaptation value of the user's dynamic learning portrait based on the communication topic in the user dialogue information.

6. The intelligent management method for foreign language scenario teaching based on virtual reality according to claim 1, characterized in that: In the process of dynamically updating the scenario teaching information selection set based on the virtual scenario model in step S400, The user's teaching bias based on the virtual scenario model is obtained, and the knowledge points whose intersection between the corresponding bound communication topic summary set and the user's teaching bias based on the virtual scenario model in the scenario teaching information alternative set based on the virtual scenario model is empty are eliminated to obtain the updated result of the scenario teaching information alternative set based on the virtual scenario model, and the elements of the communication topic summary set bound to each element in the updated scenario teaching information alternative set based on the virtual scenario model all belong to the user's teaching bias based on the virtual scenario model.

7. The intelligent management system for foreign language situational teaching based on virtual reality applies the method for intelligent management of foreign language situational teaching based on virtual reality according to any one of claims 1-6, and is characterized in that: The system includes a virtual scenario model building module, a scenario group learning heat distribution analysis module, a dynamic learning portrait generation module and a scenario teaching content management module; The virtual scenario model building module extracts keywords from the human-computer interaction information during the scenario construction phase, matches the dialogue scenarios and virtual characters in the database, and generates a virtual scenario model. The scenario group learning heat distribution analysis module retrieves a set of scenario teaching information options based on the virtual scenario model based on the scenario group learning heat distribution map in the historical data. The dynamic learning portrait generation module obtains user dialogue information in real time during the scenario teaching process, extracts the user's semantic features based on language recognition technology, and extracts the user's language expression features based on eye tracking technology, to generate a dynamic learning portrait of the user. The scenario teaching content management module generates the user's teaching preference based on the virtual scenario model based on the user's dynamic learning portrait; Combined with the user's teaching preference based on the virtual scenario model, the alternative set of scenario teaching information based on the virtual scenario model is dynamically updated.

8. The intelligent management system for foreign language situational teaching based on virtual reality according to claim 7, wherein: The dynamic learning portrait generation module includes a semantic feature extraction unit, a language expression feature extraction unit and a portrait construction unit. The semantic feature extraction unit obtains user conversation information in real time during the situational teaching process and extracts the user's semantic features based on language recognition technology; The language expression feature extraction unit obtains user dialogue information in real time during the situational teaching process and extracts the user's language expression features based on eye tracking technology; The portrait construction unit generates a dynamic learning portrait of the user based on the acquisition results of the semantic feature extraction unit and the language expression feature extraction unit.

9. The intelligent management system for foreign language scenario teaching based on virtual reality according to claim 7, characterized in that: The scenario teaching content management module includes a bias analysis unit and a data update unit. The bias analysis unit generates a teaching bias of the user based on the virtual scenario model according to the user's dynamic learning profile; The data update unit dynamically updates the alternative set of scenario teaching information based on the virtual scenario model in combination with the teaching bias of the user based on the virtual scenario model.

Images