Interactive teaching strategy optimization method

Abstract

The invention discloses an interactive teaching strategy optimization method, and relates to the technical field of interactive teaching, and the method comprises the following steps: collecting desensitization real case resources and full-link teaching behavior data of a target teaching field, and constructing a target teaching standard; the target teaching standard comprises a case scene feature area, an interactive behavior diagnosis area and a strategy intelligent iteration area; wherein the case scene feature area comprises case scene element feature factors and teaching strategy adaptive weight coefficients, and the interactive behavior diagnosis area comprises multi-modal interactive anomaly feature factors and teaching strategy defect diagnosis factors; case deduction interaction data, trainee behavior track data and knowledge mastering state data in a target teaching scene are collected, so that a personalized strategy optimization scheme can be generated according to operation habits and knowledge weak links of different trainees, one-person-one-strategy accurate training is realized, and the problem of weak point repetition caused by unified teaching is avoided.

Description

Technical Field

[0001] This invention relates to the field of interactive teaching technology, and more specifically, to a method for optimizing interactive teaching strategies. Background Technology

[0002] In vocational skills and professional field teaching, the one-way output and static solidification of traditional teaching strategies are no longer suitable for the dynamic needs of modern teaching. This is especially true in fields with strong practical application, such as police response and emergency handling, where the disconnect between teaching and practice is becoming increasingly prominent.

[0003] Taking emergency response training as an example, past training relied heavily on offline simulations and fixed case exercises: the case database has a long update cycle and often lags behind the complex changes in actual police situations, resulting in a disconnect between the scenarios trainees encounter and the real law enforcement environment; the teaching strategy is promoted with a uniform standard, which cannot be personalized according to the different trainees' behavioral patterns, such as their operational habits of information collection when receiving a call, and their knowledge mastery status, such as the error-prone links in the handling process. Some trainees repeatedly show weaknesses in key links, but it is difficult to receive accurate training and guidance.

[0004] Meanwhile, traditional teaching lacks effective use of process data. Data such as case study interaction records and student operation trajectories generated during teaching are mostly archived as training files and are not transformed into a basis for strategy optimization. Abnormal behaviors in multimodal interaction, such as delayed voice response and violation of operation steps, cannot be linked to the defects of teaching strategies, resulting in strategy adjustments relying on human experience summaries, which have problems of delayed feedback and insufficient pertinence. Summary of the Invention

[0005] In view of the shortcomings of existing technologies, the purpose of this invention is to provide an interactive teaching strategy optimization method.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An interactive teaching strategy optimization method, the method comprising the following steps:

[0008] Collect anonymized real-world case resources and full-chain teaching behavior data in the target teaching domain to construct target teaching standards; the target teaching standards include a case scenario feature area, an interactive behavior diagnosis area, and a strategy intelligent iteration area; wherein, the case scenario feature area includes case scenario element feature factors and teaching strategy adaptation weight coefficients, and the interactive behavior diagnosis area includes multimodal interaction abnormality feature factors and teaching strategy defect diagnosis factors;

[0009] Collect case study interaction data, student behavior trajectory data, and knowledge mastery status data in the target teaching scenario, and combine them with the target teaching standards to conduct feature matching to obtain the real-time effectiveness evaluation index of the target teaching strategy;

[0010] The abnormal interaction feature factors extracted from the target teaching scenario are matched with the preset multimodal interaction abnormal feature factors to obtain the corresponding teaching strategy defect diagnosis factors. Based on the teaching strategy defect diagnosis factors, the targeted optimization dimension of the current teaching strategy is determined.

[0011] The real-time performance evaluation index and targeted optimization dimensions are transmitted to the strategy intelligent iteration area to generate personalized teaching strategy optimization solutions for the target teaching scenario.

[0012] Preferably, de-identified real-world case resources and full-chain teaching behavior data are collected in the target teaching domain to construct target teaching standards, specifically including the following steps:

[0013] Collect original case information from different teaching stages. The original case information includes identity-related information and privacy-related information.

[0014] The core scenario elements, teaching implementation process, and interactive feedback results of cases in different teaching stages are obtained to generate a set of teaching case information.

[0015] By integrating native case information with teaching case information sets, de-identified real case resources are generated.

[0016] Collect target teaching data units, interactive data units, and learning data units to generate a teaching data unit library, and extract features from the teaching data unit library to form a full-link behavioral data set;

[0017] Based on anonymized real-world case resources and full-chain teaching behavior data, we construct target teaching standards.

[0018] Preferably, the following steps are taken: Data on case study interactions, student behavior patterns, and knowledge acquisition status are collected within the target teaching scenario. This data is then combined with target teaching standards to perform feature matching and obtain a real-time effectiveness evaluation index for the target teaching strategy.

[0019] Multi-dimensional feature extraction and fusion of case study interaction data, student behavior trajectory data, and knowledge mastery status data are performed to generate a set of feature factors for monitoring the teaching process.

[0020] The set of teaching process monitoring feature factors is input into the target teaching standard and matched with the case scenario element feature factors in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than the preset threshold are selected.

[0021] The strategy deviation impact index is obtained by calculating the strategy deviation characteristic factor based on the teaching strategy adaptation weight coefficient, and the real-time effectiveness evaluation index of the target teaching strategy is obtained by combining it with the preset teaching effectiveness benchmark value.

[0022] Preferably, the case scenario feature region includes case scenario element feature factors and teaching strategy adaptation weight coefficients, specifically including the following steps:

[0023] The desensitized real-world case resources in the target teaching area are decomposed into scene elements, knowledge point associations are mined, and teaching key points and difficulties are identified to generate corresponding case scene element feature factors.

[0024] Based on the full-link teaching behavior data, the influence weight of the characteristic factors of each case scenario on teaching effectiveness is calculated, and the corresponding teaching strategy adaptation weight coefficient is obtained.

[0025] Preferably, the interactive behavior diagnostic area includes multimodal interaction anomaly feature factors and teaching strategy defect diagnostic factors, specifically including the following steps:

[0026] Collect historical interaction anomaly data in the target teaching scenario, including multimodal interaction behavior anomaly monitoring records and teaching strategy defect diagnosis reports;

[0027] Feature extraction is performed on the monitoring records of multimodal interaction behavior anomalies to generate multimodal interaction anomaly feature factors;

[0028] Feature extraction is performed on the diagnostic report of teaching strategy defects to generate diagnostic factors for teaching strategy defects.

[0029] Preferably, the case study interaction data, student behavior trajectory data, and knowledge mastery status data are subjected to multi-dimensional feature extraction and fusion to generate a set of feature factors for monitoring the teaching process. Specifically, this includes the following steps:

[0030] Perform logic compliance assessment and interaction depth evaluation on case simulation interaction data, and extract case simulation interaction feature parameters;

[0031] Perform behavioral pattern recognition and operational compliance judgment on student behavior trajectory data, and extract student behavior feature parameters;

[0032] The system analyzes the knowledge mastery status data to statistically determine the accuracy of knowledge points, pinpoint weak knowledge areas, and extract characteristic parameters of knowledge mastery.

[0033] The teaching process monitoring feature factors are generated by integrating case study interaction feature parameters, student behavior feature parameters, and knowledge mastery feature parameters.

[0034] Preferably, the set of teaching process monitoring feature factors is input into the target teaching standard and matched with the case scenario element feature factors in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than a preset threshold are then selected. Specifically, this includes the following steps:

[0035] The characteristic parameters of the set of monitoring characteristic factors of the teaching process are compared with the threshold range of the standard characteristic parameters of the characteristic factors of the case scenario elements;

[0036] If the characteristic parameters of the teaching process monitoring feature factors exceed the standard feature parameter threshold range, they are marked as candidate deviation feature factors;

[0037] Calculate the deviation of the characteristic parameters of the candidate deviation feature factors from the center value of the threshold range of the standard feature parameters, and screen out the candidate deviation feature factors whose deviation is higher than the preset threshold, and determine them as strategy deviation feature factors.

[0038] Preferably, the strategy deviation impact index is calculated based on the teaching strategy adaptation weight coefficient to obtain the strategy deviation characteristic factor, and then combined with the preset teaching effectiveness benchmark value to calculate the real-time effectiveness evaluation index of the target teaching strategy. This specifically includes the following steps:

[0039] The deviation magnitude of each strategy deviation characteristic factor is combined with the corresponding teaching strategy adaptation weight coefficient to obtain the strategy deviation impact index.

[0040] The real-time effectiveness evaluation index of the target teaching strategy is obtained by calculating the teaching effectiveness benchmark value and the strategy deviation impact index.

[0041] Preferably, the abnormal interaction feature factors extracted from the target teaching scenario are matched with preset multimodal interaction abnormal feature factors to obtain corresponding teaching strategy defect diagnostic factors, specifically including the following steps:

[0042] Extract abnormal interaction feature factors from the collected data to generate a corresponding set of abnormal interaction feature parameters;

[0043] Similar feature parameters are obtained by matching the set of abnormal interaction feature parameters with the set of abnormal feature parameters of multimodal interaction anomaly feature factors.

[0044] Diagnostic factors for teaching strategy defects are based on matching similar feature parameters.

[0045] Compared with existing technologies, this invention has the following beneficial effects: By collecting anonymized real-world case resources in the target teaching field and constructing case scenario feature areas, real-time synchronization between cases and actual scenarios is achieved. Anonymized real-world police incidents can be directly transformed into teaching cases, and the core scenario elements of the cases are linked to the teaching strategy adaptation weight coefficient, significantly improving the matching degree between the teaching scenario and the real law enforcement environment, and ensuring that the training content encountered by trainees meets actual needs. By collecting trainee behavior trajectory data and knowledge mastery status data, and combining them with target teaching standards to conduct feature matching, personalized strategy optimization plans can be generated for different trainees' operating habits and knowledge weaknesses, achieving precise training for each individual and avoiding the problem of repeated weaknesses caused by uniform teaching. By transforming the full-link teaching behavior data, including case simulation interaction data and multimodal interaction anomaly data, into the core basis for strategy optimization, and by extracting abnormal interaction feature factors and matching them with teaching strategy defect diagnostic factors, a direct correlation is established between the process data generated in teaching and strategy defects, making the feedback of strategy optimization more timely and targeted. Attached Figure Description

[0046] Fig. 1 This is a schematic diagram illustrating the steps of an interactive teaching strategy optimization method provided in an embodiment of the present invention;

[0047] Fig. 2 This is a schematic diagram illustrating the steps of constructing target teaching standards in an interactive teaching strategy optimization method provided by an embodiment of the present invention. Detailed Implementation

[0048] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0049] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0051] Reference Figs. 1-2 As shown.

[0052] This embodiment further illustrates the interactive teaching strategy optimization method proposed in this invention.

[0053] An interactive teaching strategy optimization method, the method comprising the following steps:

[0054] Collect anonymized real-world case resources and full-chain teaching behavior data in the target teaching domain to construct target teaching standards. The target teaching standards include a case scenario feature area, an interactive behavior diagnosis area, and a strategy intelligent iteration area. The case scenario feature area includes case scenario element feature factors and teaching strategy adaptation weight coefficients, and the interactive behavior diagnosis area includes multimodal interaction abnormality feature factors and teaching strategy defect diagnosis factors.

[0055] Collect case study interaction data, student behavior trajectory data, and knowledge mastery status data in the target teaching scenario, and combine them with the target teaching standards to conduct feature matching to obtain the real-time effectiveness evaluation index of the target teaching strategy;

[0056] The abnormal interaction feature factors extracted from the target teaching scenario are matched with the preset multimodal interaction abnormal feature factors to obtain the corresponding teaching strategy defect diagnosis factors. Based on the teaching strategy defect diagnosis factors, the targeted optimization dimension of the current teaching strategy is determined.

[0057] The real-time performance evaluation index and targeted optimization dimensions are transmitted to the strategy intelligent iteration area to generate personalized teaching strategy optimization solutions for the target teaching scenario.

[0058] Specifically, once the real-time performance evaluation index and targeted optimization dimension in the target teaching scenario are determined, they will be synchronously transmitted to the strategy intelligent iteration area. The algorithm logic and data model preset in the strategy intelligent iteration area, combined with the actual needs of the target teaching scenario, will generate a personalized teaching strategy optimization scheme that is adapted to the scenario.

[0059] For example, in the teaching scenario of this system, the real-time performance evaluation index shows that the completeness of information collection by trainees during the call-responding inquiry process is only 60%, while the targeted optimization dimension is identified as the guiding logic of the call-responding script. At this point, the real-time performance evaluation index and the targeted optimization dimension are transmitted to the strategy intelligent iteration area. This area combines the characteristics of real-world call-responding scenarios with the teaching strategy adaptation weight coefficients to generate corresponding personalized optimization solutions. These solutions might include adjusting the questioning order in the call-responding script, adding key information prompts, and designing multiple rounds of human-computer dialogue practice to strengthen training in this area, thereby improving the efficiency and completeness of information collection by trainees during the call-responding inquiry process. The strategy intelligent iteration area, based on the real-time performance evaluation index and guided by the targeted optimization dimension, combines various factors in the target teaching standards to transform abstract evaluation data into concrete and executable teaching strategy adjustment plans. This achieves precise adaptation of the teaching strategy to the target teaching scenario, improving the relevance and effectiveness of teaching.

[0060] Collect anonymized real-world case resources and full-chain teaching behavior data in the target teaching domain, and construct target teaching standards. This includes the following steps:

[0061] Collect original case information from different teaching stages. The original case information includes identity-related information and privacy-related information.

[0062] The core scenario elements, teaching implementation process, and interactive feedback results of cases in different teaching stages are obtained to generate a set of teaching case information.

[0063] By integrating native case information with teaching case information sets, de-identified real case resources are generated.

[0064] Collect target teaching data units, interactive data units, and learning data units to generate a teaching data unit library, and extract features from the teaching data unit library to form a full-link behavioral data set;

[0065] Based on anonymized real-world case resources and full-chain teaching behavior data, we construct target teaching standards.

[0066] Collect original case information from different teaching stages. This original case information includes identity-related information and privacy-related information. Taking emergency response training as an example, the original case information may involve the basic identity of the person reporting the incident and privacy details of the reporting scenario. After collection, this information needs to be processed to remove sensitive content and ensure that the case resources meet the anonymization requirements.

[0067] The core scenario elements, teaching implementation process, and interactive feedback results of cases from different teaching stages are obtained to generate a teaching case information set. In emergency response training, the core scenario element of a case might be the degree of emotional conflict at the scene of a domestic dispute, the teaching implementation process might be the steps of receiving the call and questioning, issuing instructions, and handling the situation on-site, and the interactive feedback results might be the degree of dialogue matching between the trainees and the virtual caller in the simulated handling, and the accuracy rate of completing the handling steps. Integrating these contents forms the corresponding teaching case information set.

[0068] By integrating the collected original case information with the generated teaching case information set, de-identified real-world case resources are obtained. For example, the de-identified original police report of a family dispute is combined with the corresponding teaching process and interactive feedback data to form de-identified case resources that can be used for teaching and training, which retains the scene characteristics of the real police report while avoiding privacy leaks.

[0069] The process involves collecting target teaching data units, interaction data units, and learning data units to generate a teaching data unit library. Feature extraction is then performed to create a complete set of behavioral data. For example, in an emergency response training scenario, the target teaching data unit might be the completeness of information collection during the emergency response phase; the interaction data unit might be the frequency of dialogue between the trainee and the training system, and the operation response time; and the learning data unit might be the trainee's assessment scores under different emergency types. After integrating this data into the teaching data unit library, key features are extracted, such as the fluctuation range of information collection completeness and the correlation between dialogue frequency and response efficiency, forming complete teaching behavioral data.

[0070] Based on the compiled de-identified real-world case resources and full-link teaching behavior data, target teaching standards are constructed. For example, by combining the scenario characteristics of de-identified family dispute cases with the full-link behavior data of students in such cases, the appropriate weight of the emergency call response guidance strategy is determined. At the same time, the strategy defect factors corresponding to abnormal characteristics such as insufficient dialogue interaction are extracted. Finally, a complete optimization system covering cases, interactions, and iterations is built to support the adjustment of subsequent teaching strategies.

[0071] Collect case study interaction data, student behavior trajectory data, and knowledge mastery status data in the target teaching scenario, and combine them with target teaching standards to conduct feature matching to obtain a real-time effectiveness evaluation index for the target teaching strategy. The specific steps include:

[0072] Multi-dimensional feature extraction and fusion of case study interaction data, student behavior trajectory data, and knowledge mastery status data are performed to generate a set of feature factors for monitoring the teaching process.

[0073] The set of teaching process monitoring feature factors is input into the target teaching standard and matched with the case scenario element feature factors in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than the preset threshold are selected.

[0074] The strategy deviation impact index is obtained by calculating the strategy deviation characteristic factor based on the teaching strategy adaptation weight coefficient, and the real-time effectiveness evaluation index of the target teaching strategy is obtained by combining it with the preset teaching effectiveness benchmark value.

[0075] Data on case simulation interactions, student behavior trajectories, and knowledge acquisition status are collected from target teaching scenarios. Multi-dimensional feature extraction and fusion are then performed on these data to generate a set of monitoring feature factors for the teaching process. For example, in emergency response training, case simulation interaction data includes the duration of dialogue between students and virtual callers, and the number of times key information such as location and personnel details are requested. Student behavior trajectories include the duration of student stays at each stage of the simulated response process and the frequency of jumps in operational steps. Knowledge acquisition status data includes students' assessment scores on the procedures for handling dispute-related police incidents and the recurrence rate of errors in error-prone steps. After extracting features from these data, such as the sufficiency of key information requests, the standardization of operational steps, and the degree of achievement of assessment scores, these features are fused to form a set of monitoring feature factors for the teaching process.

[0076] The set of monitoring feature factors for the teaching process is input into the target teaching standards and matched with the case scenario element feature factors in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than a preset threshold are then selected. For example, in handling domestic dispute police reports, the case scenario element feature factors in the case scenario feature area include: the priority of the emotional reassurance aspect of the incident and the mandatory fields for key information collection. If the matching degree between the percentage of emotional reassurance operations and the completeness of key information collection in the set of monitoring feature factors for the teaching process is lower than a preset threshold of 80%, then these two items will be identified as strategy deviation feature factors.

[0077] The strategy deviation impact index is calculated based on the teaching strategy adaptation weight coefficient to obtain the strategy deviation characteristic factor. Combined with the preset teaching effectiveness benchmark value, the real-time effectiveness evaluation index of the target teaching strategy is calculated. Assuming that in the teaching strategy adaptation weight coefficient, the weight of the emotional soothing step is 0.4, and the weight of key information collection is 0.6; and the deviation degrees corresponding to the strategy deviation characteristic factors are 20% for the operational proportion deviation of the emotional soothing step and 30% for the completeness deviation of key information collection, then the formula for calculating the strategy deviation impact index is: Strategy Deviation Impact Index = (Degree of Deviation in Emotional Soothing Step × Corresponding Weight) + (Degree of Deviation in Key Information Collection × Corresponding Weight). After substituting the values, the Strategy Deviation Impact Index = (20% × 0.4) + (30% × 0.6) = 26%. If the preset teaching effectiveness benchmark value is 100 points, then the formula for calculating the real-time effectiveness evaluation index is: real-time effectiveness evaluation index = teaching effectiveness benchmark value × (1 - strategy deviation impact index), that is, 100 × (1 - 26%) = 74 points. This score is the real-time effectiveness evaluation result of the teaching strategy for handling family dispute police incidents.

[0078] The case scenario feature area includes case scenario element feature factors and teaching strategy adaptation weight coefficients, specifically including the following steps:

[0079] The desensitized real-world case resources in the target teaching area are decomposed into scene elements, knowledge point associations are mined, and teaching key points and difficulties are identified to generate corresponding case scene element feature factors.

[0080] Based on the full-link teaching behavior data, the influence weight of the characteristic factors of each case scenario on teaching effectiveness is calculated, and the corresponding teaching strategy adaptation weight coefficient is obtained.

[0081] The process involves desensitizing real-world case resources for the target teaching area, breaking down scene elements, mining knowledge point associations, and identifying key teaching points and difficulties. This generates corresponding case scene element feature factors. In emergency response training, the desensitized real-world case resources are desensitized cases of shop disputes. Scene element desensitization extracts elements such as the intensity of the caller's emotions, the number of people on site, and the amount of money involved in the dispute. Knowledge point association mining correlates these elements with teaching knowledge points such as emergency response protocol standards and key points for on-site order control. Identifying key teaching points and difficulties identifies emotional calming techniques and the priority of handling multi-person scenarios as key teaching points for this case. Integrating these elements forms the case scene element feature factors corresponding to shop dispute cases.

[0082] Based on end-to-end teaching behavior data, the influence weights of characteristic factors of each case scenario element on teaching effectiveness are calculated, resulting in corresponding teaching strategy adaptation weight coefficients. In emergency response training, end-to-end teaching behavior data includes data on the completeness of emergency information collection, the quality of emotional reassurance, and the compliance of on-site handling procedures during training for this type of emergency. Taking the characteristic factor of emotional reassurance communication skills as an example, by analyzing the fluctuation of student assessment scores corresponding to this element in the end-to-end data—for instance, when students master this skill, their assessment scores increase by an average of 25%, while the influence of other elements on scores is relatively low—the influence weight of this element on teaching effectiveness can be calculated.

[0083] Assuming the impact weight of each case scenario element characteristic factor on teaching effectiveness is calculated through the proportion of element contribution, the formula is: Weight of case scenario element characteristic factor = Teaching effectiveness improvement value corresponding to the element ÷ Sum of teaching effectiveness improvement values ​​of all elements. Taking a shop dispute police incident as an example, if the teaching effectiveness improvement value corresponding to emotional calming techniques is 25, the improvement value corresponding to the handling priority in multi-person scenarios is 20, and the sum of the improvement values ​​of other elements is 35, then the weight of emotional calming techniques = 25 ÷ (25 + 20 + 35) = 0.3125, and the weight of the handling priority in multi-person scenarios = 20 ÷ 80 = 0.25. These calculation results are the teaching strategy adaptation weight coefficients corresponding to the teaching scenario element characteristic factors of each case scenario.

[0084] The interactive behavior diagnostic area includes multimodal interaction anomaly characteristic factors and teaching strategy defect diagnostic factors, specifically including the following steps:

[0085] Collect historical interaction anomaly data in the target teaching scenario, including multimodal interaction behavior anomaly monitoring records and teaching strategy defect diagnosis reports;

[0086] Feature extraction is performed on the monitoring records of multimodal interaction behavior anomalies to generate multimodal interaction anomaly feature factors;

[0087] Feature extraction is performed on the diagnostic report of teaching strategy defects to generate diagnostic factors for teaching strategy defects.

[0088] Historical interaction anomaly data was collected for the target teaching scenarios. This data included multimodal interaction behavior anomaly monitoring records and teaching strategy defect diagnosis reports. In emergency response training, multimodal interaction behavior anomaly monitoring records included entries where trainees' voice response intervals exceeded 10 seconds during simulated emergency response, and operation logs showing more than three errors when entering key information in text. The teaching strategy defect diagnosis reports addressed these anomalies, summarizing past findings such as missing key prompts in the emergency response script guidance module and unreasonable layout of the information entry interface.

[0089] Feature extraction is performed on monitoring records of abnormal multimodal interaction behaviors to generate multimodal interaction anomaly feature factors. Taking the abnormal voice response interval in emergency call handling training as an example, features such as response interval duration, the teaching segment in which the abnormality occurred, and the type of emergency call in which the abnormality occurred are extracted from the monitoring records. For example, in the emergency call handling process for family disputes, the student's voice response interval repeatedly exceeded 10 seconds. After integrating these features, a multimodal interaction anomaly feature factor of delayed voice response in the emergency call handling process for family disputes is formed. Similarly, from records of information entry errors, features such as error type and error frequency can be extracted to generate a multimodal interaction anomaly feature factor of insufficient accuracy in key information entry.

[0090] Feature extraction is performed on teaching strategy defect diagnosis reports to generate teaching strategy defect diagnostic factors. For reports of missing key prompts in the call handling script guidance module during call handling training, features of the corresponding teaching module and the abnormal behavior types caused by the defect are extracted to form a teaching strategy defect diagnostic factor for missing information in the call handling script guidance module. For reports of unreasonable layout of information input interface, features of the corresponding interactive interface and the operational efficiency affected by the defect are extracted to generate a teaching strategy defect diagnostic factor for insufficient adaptability of information input interface layout.

[0091] The extracted multimodal interaction anomaly feature factors are correlated with teaching strategy defect diagnostic factors. For example, the anomaly feature factor of delayed voice response in the call-receiving process for family dispute-related police incidents is matched with the defect diagnostic factor of missing information in the call-receiving script guidance module, providing accurate diagnostic basis for subsequent optimization of teaching strategies.

[0092] The process of extracting and fusing multi-dimensional features from case study interaction data, student behavior trajectory data, and knowledge mastery status data to generate a set of feature factors for monitoring the teaching process includes the following steps:

[0093] Perform logic compliance assessment and interaction depth evaluation on case simulation interaction data, and extract case simulation interaction feature parameters;

[0094] Perform behavioral pattern recognition and operational compliance judgment on student behavior trajectory data, and extract student behavior feature parameters;

[0095] The system analyzes the knowledge mastery status data to statistically determine the accuracy of knowledge points, pinpoint weak knowledge areas, and extract characteristic parameters of knowledge mastery.

[0096] The teaching process monitoring feature factors are generated by integrating case study interaction feature parameters, student behavior feature parameters, and knowledge mastery feature parameters.

[0097] The interactive data from case simulations is used to assess the compliance of the simulation logic and the depth of interaction, extracting characteristic parameters of the case simulation interaction. In emergency response training, case simulation interactive data consists of the interaction records of trainees simulating handling emergency situations, such as simulation data of handling domestic disputes. The compliance assessment of the simulation logic checks whether trainees follow the standardized procedures of emergency response inquiry, instruction reporting, and on-site reassurance. If a trainee skips the instruction reporting and directly handles the situation, it is considered a logical non-compliance. The depth of interaction assessment counts the number of effective dialogue rounds between the trainee and the virtual caller. For example, if the number of dialogue rounds is less than 5, it is considered insufficient interaction. Through these judgments and assessments, characteristic parameters of case simulation interaction, such as the compliance rate of the simulation process and the percentage of effective interaction rounds, are extracted. For example, the trainee's simulation process compliance rate is 80%, and the percentage of effective interaction rounds is 60%.

[0098] The system performs behavioral pattern recognition and operational compliance assessment on trainee behavior trajectory data, extracting trainee behavioral characteristic parameters. Trainee behavior trajectory data consists of trainees' operational records in the training system, such as the steps for entering alarm information and the response time for sending handling instructions. Behavioral pattern recognition analyzes trainees' operating habits, such as whether there is a reasonable pattern of entering location information first and then inquiring about personnel. Operational compliance assessment verifies whether operations conform to regulations; for example, failure to confirm information before sending instructions is considered a violation. Based on these analyses, trainee behavioral characteristic parameters such as the degree of standardization of operational steps and the compliance rate of response time are extracted. For example, a trainee's operational step standardization rate is 75%, and the compliance rate of response time is 85%.

[0099] The system analyzes knowledge mastery data to statistically determine the accuracy rate of knowledge points, identify weaknesses, and extract characteristic parameters. Knowledge mastery data represents trainees' assessment and practice results, such as their answers to questions on emergency response techniques and their scores on handling procedures. The accuracy rate calculation measures the percentage of correct answers for each knowledge point; for example, a 70% accuracy rate for calming trainees in dispute-related emergency situations. Weakness identification identifies knowledge points with accuracy rates below a preset threshold, such as 80%. For instance, a 60% accuracy rate for prioritizing handling multiple personnel in a situation is considered a weak point. From this, the system extracts characteristic parameters such as the average accuracy rate of knowledge points and the percentage of weak points. For example, a trainee's average accuracy rate for knowledge points might be 72%, with weak points accounting for 20%.

[0100] By integrating interactive feature parameters from case simulations, behavioral feature parameters, and knowledge mastery feature parameters, a teaching process monitoring feature factor is generated. Taking a practical training session on family disputes in emergency response training as an example, the parameters of the trainee's simulation process compliance rate (80%), operational step standardization (75%), and average knowledge point accuracy rate (72%) are integrated to generate a teaching process monitoring feature factor with a comprehensive standardization rate of 75.7% for the family dispute emergency training. If a weighted average calculation is used, the formula is: Comprehensive Standardization Rate = Simulation Process Compliance Rate × 0.3 + Operational Step Standardization × 0.3 + Average Knowledge Point Accuracy Rate × 0.4. Substituting the values, we get 80% × 0.3 + 75% × 0.3 + 72% × 0.4 = 75.3%, forming a complete set of teaching process monitoring feature factors.

[0101] The set of monitoring feature factors of the teaching process is input into the target teaching standard and matched with the feature factors of case scenario elements in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than a preset threshold are screened out. The specific steps include:

[0102] The characteristic parameters of the set of monitoring characteristic factors of the teaching process are compared with the threshold range of the standard characteristic parameters of the characteristic factors of the case scenario elements;

[0103] If the characteristic parameters of the teaching process monitoring feature factors exceed the standard feature parameter threshold range, they are marked as candidate deviation feature factors;

[0104] Calculate the deviation of the characteristic parameters of the candidate deviation feature factors from the center value of the threshold range of the standard feature parameters, and screen out the candidate deviation feature factors whose deviation is higher than the preset threshold, and determine them as strategy deviation feature factors.

[0105] The characteristic parameters of the teaching process monitoring feature factor set are compared with the standard feature parameter threshold ranges of the case scenario element feature factors. In this scenario, the case scenario element feature factors include the completeness of key information collection upon receiving the alarm and the compliance rate of on-site handling steps, with the corresponding standard feature parameter threshold ranges set to 80% to 90% and 85% to 95%, respectively. In the teaching process monitoring feature factor set, the completeness parameter of key information collection upon receiving the alarm for a certain batch of students is 65%, and the compliance rate parameter of on-site handling steps is 70%.

[0106] If the characteristic parameters of the teaching process monitoring feature factors exceed the standard feature parameter threshold range, they are marked as candidate deviation feature factors. For example, if the completeness of the student's alarm key information collection is 65%, which is lower than the standard threshold lower limit of 80%, and the compliance rate of the on-site handling steps is 70%, which is lower than the standard threshold lower limit of 85%, the factors corresponding to these two feature parameters are marked as candidate deviation feature factors.

[0107] The deviation of the characteristic parameters of candidate deviation feature factors from the central value of the threshold range of standard feature parameters is calculated. Candidate deviation feature factors with deviations exceeding a preset threshold are selected and identified as strategy deviation feature factors. For example, if the preset deviation threshold is 15%, the central value of the threshold range of each standard feature parameter is first calculated: the central value of the completeness of alarm key information collection is (80%+90%)÷2=85%, and the central value of the compliance rate of on-site handling steps is (85%+95%)÷2=90%. The formula for calculating the deviation is: Where PF represents the deviation magnitude, HC represents the candidate parameter, and ZH represents the center value. After substituting the values, the deviation magnitude of the completeness of the alarm key information collection is: |(65%-85%)÷85%|×100%≈23.5%, and the deviation magnitude of the compliance rate of on-site handling steps is |(70%-90%)÷90%|×100%≈22.2%. The deviation magnitudes of these two candidate deviation feature factors are both higher than the preset threshold of 15%, so they will be identified as strategy deviation feature factors, providing precise guidance for the optimization of subsequent teaching strategies.

[0108] The strategy deviation impact index is calculated based on the teaching strategy adaptation weight coefficient to determine the strategy deviation characteristic factor. This is then combined with a pre-set teaching effectiveness benchmark value to calculate the real-time effectiveness evaluation index of the target teaching strategy. The specific steps include:

[0109] The deviation magnitude of each strategy deviation characteristic factor is combined with the corresponding teaching strategy adaptation weight coefficient to obtain the strategy deviation impact index.

[0110] The real-time effectiveness evaluation index of the target teaching strategy is obtained by calculating the teaching effectiveness benchmark value and the strategy deviation impact index.

[0111] The deviation magnitudes of each strategy deviation characteristic factor are fused with the corresponding teaching strategy adaptation weight coefficients to obtain the strategy deviation impact index. In this scenario, the identified strategy deviation characteristic factors include the completeness of key alarm information collection and the compliance rate of on-site handling procedures. The corresponding teaching strategy adaptation weight coefficients are set to 0.4 and 0.6, respectively, with a total weight coefficient of 1, representing the degree of influence of different factors on the teaching strategy. The calculated deviation magnitudes are 23.5% and 22.2%, respectively. The fusion calculation formula is as follows: Where SDEI is the policy bias impact index, and D i W represents the deviation magnitude of the i-th strategy deviation characteristic factor. i The teaching strategy adaptation weight coefficient is set for the i-th strategy deviation feature factor. After substituting the values, we get 23.5%×0.4+22.2%×0.6=9.4%+13.32%=22.72%, which is the strategy deviation impact index in this teaching scenario.

[0112] Next, the teaching effectiveness benchmark value and the strategy deviation impact index are calculated to obtain the real-time effectiveness evaluation index of the target teaching strategy. Assuming a preset teaching effectiveness benchmark value of 100 points, representing the effectiveness when the teaching strategy fully meets the standard, the calculation formula is: Real-time effectiveness evaluation index = Teaching effectiveness benchmark value × (1 - Strategy deviation impact index). Substituting the values, we get 100 × (1 - 22.72%) = 77.28 points. This score represents the real-time effectiveness evaluation result of the current neighborhood conflict-related police incident teaching strategy. It directly reflects the reduced effectiveness level of the strategy due to deviations, clearly indicating the impact of teaching strategy deviations on overall effectiveness and providing a quantitative basis for subsequent strategy adjustments. For example, a real-time effectiveness evaluation index of 77.28 points in this scenario indicates that the strategies corresponding to the key information collection and on-site handling steps need to be optimized to improve teaching effectiveness.

[0113] The abnormal interaction feature factors extracted from the target teaching scenario are matched with preset multimodal interaction abnormal feature factors to obtain corresponding teaching strategy defect diagnostic factors. The specific steps include:

[0114] Extract abnormal interaction feature factors from the collected data to generate a corresponding set of abnormal interaction feature parameters;

[0115] Similar feature parameters are obtained by matching the set of abnormal interaction feature parameters with the set of abnormal feature parameters of multimodal interaction anomaly feature factors.

[0116] Diagnostic factors for teaching strategy defects are based on matching similar feature parameters.

[0117] Anomaly interaction feature factors are extracted from the collected data to generate a corresponding set of anomaly interaction feature parameters. In scenario-based teaching and training, the collected data includes records of interactions between trainees and the system, such as abnormal behaviors like an interval exceeding 15 seconds when asking for key information via voice during an alarm call, or more than two errors when entering accident location information. Feature factors are extracted from this data, such as voice inquiry response delay and information entry error frequency, and the corresponding parameters are recorded: average voice inquiry interval of 18 seconds, and information entry error frequency of 2 times / case, thereby generating a set of anomaly interaction feature parameters.

[0118] Next, the set of abnormal interaction feature parameters is matched with the set of abnormal feature parameters of multimodal interaction abnormal feature factors to obtain similar feature parameters. The preset multimodal interaction abnormal feature factors include a set of parameters corresponding to voice response delay anomalies (interview interval ≥ 12 seconds) and information entry error anomalies (error frequency ≥ 1 per case). By comparing the currently extracted voice interview interval of 18 seconds and information entry error frequency of 2 per case with the preset parameter set, similar feature parameters corresponding to voice response delay anomalies and information entry error anomalies can be matched.

[0119] Finally, diagnostic factors for teaching strategy defects are matched based on similar feature parameters. Pre-defined multimodal interaction anomaly feature factors are correlated with these diagnostic factors. For example, the diagnostic factor for voice response delay anomalies is the lack of real-time prompts in the alarm response guidance module, while the diagnostic factor for information entry errors is the vague design of the accident location information entry interface options. By using the matched similar feature parameters, these two diagnostic factors for teaching strategy defects can be located, clarifying the specific aspects of the current teaching strategy that need optimization.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An interactive teaching strategy optimization method, characterized in that, The method includes the following steps: Collect anonymized real-world case resources and full-chain teaching behavior data in the target teaching domain to construct target teaching standards; the target teaching standards include a case scenario feature area, an interactive behavior diagnosis area, and a strategy intelligent iteration area; wherein, the case scenario feature area includes case scenario element feature factors and teaching strategy adaptation weight coefficients, and the interactive behavior diagnosis area includes multimodal interaction abnormality feature factors and teaching strategy defect diagnosis factors; Collect case study interaction data, student behavior trajectory data, and knowledge mastery status data in the target teaching scenario, and combine them with the target teaching standards to conduct feature matching to obtain the real-time effectiveness evaluation index of the target teaching strategy; The abnormal interaction feature factors extracted from the target teaching scenario are matched with the preset multimodal interaction abnormal feature factors to obtain the corresponding teaching strategy defect diagnosis factors. Based on the teaching strategy defect diagnosis factors, the targeted optimization dimension of the current teaching strategy is determined. The real-time performance evaluation index and targeted optimization dimensions are transmitted to the strategy intelligent iteration area to generate personalized teaching strategy optimization solutions for the target teaching scenario.

2. The interactive teaching strategy optimization method according to claim 1, characterized in that, Collect anonymized real-world case resources and full-chain teaching behavior data in the target teaching domain, and construct target teaching standards. This includes the following steps: Collect original case information from different teaching stages. The original case information includes identity-related information and privacy-related information. The core scenario elements, teaching implementation process, and interactive feedback results of cases in different teaching stages are obtained to generate a set of teaching case information. By integrating native case information with teaching case information sets, de-identified real case resources are generated. Collect target teaching data units, interactive data units, and learning data units to generate a teaching data unit library, and extract features from the teaching data unit library to form a full-link behavioral data set; Based on anonymized real-world case resources and full-chain teaching behavior data, we construct target teaching standards.

3. The interactive teaching strategy optimization method according to claim 2, characterized in that, Collect case study interaction data, student behavior trajectory data, and knowledge mastery status data in the target teaching scenario, and combine them with target teaching standards to conduct feature matching to obtain a real-time effectiveness evaluation index for the target teaching strategy. The specific steps include: Multi-dimensional feature extraction and fusion of case study interaction data, student behavior trajectory data, and knowledge mastery status data are performed to generate a set of feature factors for monitoring the teaching process. Input the set of teaching process monitoring feature factors into the target teaching standard, match it with the case scenario element feature factors in the case scenario feature area, and filter out the strategy deviation feature factors with a matching degree lower than the preset threshold. The strategy deviation impact index is obtained by calculating the strategy deviation characteristic factor based on the teaching strategy adaptation weight coefficient, and the real-time effectiveness evaluation index of the target teaching strategy is obtained by combining it with the preset teaching effectiveness benchmark value.

4. The interactive teaching strategy optimization method according to claim 3, characterized in that, The case scenario feature area includes case scenario element feature factors and teaching strategy adaptation weight coefficients, specifically including the following steps: The desensitized real-world case resources in the target teaching area are decomposed into scene elements, knowledge point associations are mined, and teaching key points and difficulties are identified to generate corresponding case scene element feature factors. Based on the full-link teaching behavior data, the influence weight of the characteristic factors of each case scenario on teaching effectiveness is calculated, and the corresponding teaching strategy adaptation weight coefficient is obtained.

5. The interactive teaching strategy optimization method according to claim 4, characterized in that, The interactive behavior diagnostic area includes multimodal interaction anomaly characteristic factors and teaching strategy defect diagnostic factors, specifically including the following steps: Collect historical interaction anomaly data in the target teaching scenario, including multimodal interaction behavior anomaly monitoring records and teaching strategy defect diagnosis reports; Feature extraction is performed on the monitoring records of multimodal interaction behavior anomalies to generate multimodal interaction anomaly feature factors; Feature extraction is performed on the diagnostic report of teaching strategy defects to generate diagnostic factors for teaching strategy defects.

6. The interactive teaching strategy optimization method according to claim 3, characterized in that, The process of extracting and fusing multi-dimensional features from case study interaction data, student behavior trajectory data, and knowledge mastery status data to generate a set of feature factors for monitoring the teaching process includes the following steps: Perform logic compliance assessment and interaction depth evaluation on case simulation interaction data, and extract case simulation interaction feature parameters; Perform behavioral pattern recognition and operational compliance judgment on student behavior trajectory data, and extract student behavior feature parameters; The system analyzes the knowledge mastery status data to statistically determine the accuracy of knowledge points, pinpoint weak knowledge areas, and extract characteristic parameters of knowledge mastery. The teaching process monitoring feature factors are generated by integrating case study interaction feature parameters, student behavior feature parameters, and knowledge mastery feature parameters.

7. The interactive teaching strategy optimization method according to claim 3, characterized in that, The set of monitoring feature factors of the teaching process is input into the target teaching standard and matched with the feature factors of case scenario elements in the case scenario feature area. Strategy deviation feature factors with a matching degree lower than a preset threshold are screened out. The specific steps include: The characteristic parameters of the set of monitoring characteristic factors of the teaching process are compared with the threshold range of the standard characteristic parameters of the characteristic factors of the case scenario elements; If the characteristic parameters of the teaching process monitoring feature factors exceed the standard feature parameter threshold range, they are marked as candidate deviation feature factors; Calculate the deviation of the characteristic parameters of the candidate deviation feature factors from the center value of the threshold range of the standard feature parameters, and screen out the candidate deviation feature factors whose deviation is higher than the preset threshold, and determine them as strategy deviation feature factors.

8. The interactive teaching strategy optimization method according to claim 3, characterized in that, The strategy deviation impact index is calculated based on the teaching strategy adaptation weight coefficient to determine the strategy deviation characteristic factor. This is then combined with a pre-set teaching effectiveness benchmark value to calculate the real-time effectiveness evaluation index of the target teaching strategy. The specific steps include: The deviation magnitude of each strategy deviation characteristic factor is combined with the corresponding teaching strategy adaptation weight coefficient to obtain the strategy deviation impact index. The real-time effectiveness evaluation index of the target teaching strategy is obtained by calculating the teaching effectiveness benchmark value and the strategy deviation impact index.

9. The interactive teaching strategy optimization method according to claim 7, characterized in that, The abnormal interaction feature factors extracted from the target teaching scenario are matched with preset multimodal interaction abnormal feature factors to obtain corresponding teaching strategy defect diagnostic factors. The specific steps include: Extract abnormal interaction feature factors from the collected data to generate a corresponding set of abnormal interaction feature parameters; Similar feature parameters are obtained by matching the set of abnormal interaction feature parameters with the set of abnormal feature parameters of multimodal interaction anomaly feature factors. Diagnostic factors for teaching strategy defects are based on matching similar feature parameters.

Images