Intelligent interaction system, method, readable storage medium and electronic device

Abstract

This invention provides an intelligent interactive system, method, readable storage medium, and electronic device with speech and behavior guidance functions. The system includes modules for data collection, cleaning, analysis, and prompt word construction, as well as a large language model invocation module. The system collects user dialogue and behavior data, extracts language and behavioral features, analyzes speech and behavior requiring improvement, and constructs prompt words containing behavior correction guidance scripts. By invoking a large language model, personalized text stories or educational content are generated. This invention solves the technical problem of how existing intelligent interactive systems can effectively monitor, analyze, and guide users to correct inappropriate speech and behavior. Through an integrated intelligent interactive system, it achieves real-time analysis and personalized feedback of user behavior, enhancing the user experience.

Description

Technical Field

[0001] This application relates to the field of intelligent interaction technology, specifically to an intelligent interaction system, method, readable storage medium, and electronic device with speech and behavior guidance function. Background Technology

[0002] In today's fast-paced society, people face various social situations and work pressures, which can lead to inappropriate words and actions. Society has an urgent need for educational tools that can promote positive behavior. With the development of artificial intelligence technology, especially the advancements in Natural Language Processing (NLP) and machine learning, intelligent interactive systems have found more application scenarios. These technologies enable systems to understand and analyze users' speech and behavior, providing personalized feedback. While intelligent interactive systems have made significant progress in information retrieval and entertainment, gaps remain in promoting improved user behavior. These systems often lack in-depth analysis of user behavior and proactive guidance strategies. Summary of the Invention

[0003] The main objective of this application is to provide an intelligent interactive system, method, readable storage medium, and electronic device with speech and behavior guidance function. It solves the technical problem in the prior art of how intelligent interactive systems can effectively monitor, analyze, and guide users to correct inappropriate speech and behavior, and realizes the technical effect of real-time analysis and personalized feedback of user behavior, thereby enhancing the user experience.

[0004] To achieve the above objectives, the present invention proposes an intelligent interactive system, comprising:

[0005] The data collection module is used to collect dialogue records between users and the intelligent interaction system, as well as user behavior data; among which, user behavior data is used to record and mark events in which users complete specific tasks or behaviors.

[0006] The data cleaning module is used to analyze the dialogue records collected by the data collection module to extract the user's language expression features, and to analyze the user behavior data collected by the data collection module to extract the user's behavior features.

[0007] The data analysis module is used to analyze the language expression features and behavioral features data generated by the data cleaning module, and to identify the speech and behavior of users that need to be improved and corrected.

[0008] The prompt word construction module is used to construct prompt words based on the undesirable speech and behaviors requiring improvement identified by the data analysis module; wherein, the prompt words include a behavior correction guidance script, and the behavior correction guidance script includes preset scenarios, roles, and dialogues; and

[0009] The large language model invocation module is used to invoke the large language model based on the prompt words constructed by the prompt word construction module to generate corresponding contextualized content; wherein, the contextualized content includes text stories or educational content.

[0010] Among them, the language expression features include, but are not limited to, vocabulary selection, emotional tone, level of politeness, language style, and communication efficiency;

[0011] The behavioral characteristics include, but are not limited to, the frequency of the behavior, its persistence, selection preferences, effect evaluation, and time series.

[0012] The data analysis module is used for:

[0013] Receive language expression features and behavioral features data processed and structured by the data cleaning module;

[0014] Machine learning algorithms are used to learn and train on historical user data to build a model for recognizing inappropriate language and behavior.

[0015] The structured language expression features and behavioral features data are input into the trained model for pattern matching and analysis to identify inappropriate speech and behavior.

[0016] Based on custom rules, assess whether user behavior falls into the category that needs improvement;

[0017] The output analysis results clearly indicate the user's inappropriate language and behavior that needs improvement.

[0018] The custom rules include frequency thresholds for the occurrence of behaviors, behavior type classifications, and behavior severity levels.

[0019] The prompts include, but are not limited to, descriptions of inappropriate language and behavior, descriptions of improvement goals, and descriptions of situational factors.

[0020] The prompt word construction module is used for:

[0021] Data preprocessing is performed on the undesirable speech and behaviors that need improvement;

[0022] Using a predefined library of undesirable behavior patterns, pattern matching technology is used to identify the parts of the undesirable speech and behavior that need improvement that conform to the characteristics of undesirable behavior;

[0023] Classify the identified misbehaviors;

[0024] Analyze the context in which the undesirable behavior occurs to determine the environment in which it takes place.

[0025] Improved suggestions are generated based on the classification results and context analysis results, and suggestion words are constructed.

[0026] The user behavior data refers to attendance records;

[0027] The data collection module is used to automatically obtain the user's check-in records by integrating with the user's check-in application.

[0028] To achieve the above objectives, the present invention also proposes an intelligent interaction method, the method comprising:

[0029] Collect dialogue records between users and intelligent interaction systems, as well as user behavior data; wherein, the user behavior data is used to record and mark events in which users complete specific tasks or behaviors;

[0030] Analyze collected dialogue records to extract users' language expression features, and analyze collected user behavior data to extract users' behavioral features;

[0031] Analyze the generated language expression and behavioral feature data to identify the speech and behavior of users that need improvement and correction;

[0032] Based on the identified inappropriate speech and behavior requiring improvement, cue words are constructed; wherein, the cue words include a behavior correction guidance script, and the behavior correction guidance script includes preset scenarios, roles, and dialogues;

[0033] The large language model is invoked based on the constructed prompt words to generate corresponding contextualized content; wherein, the contextualized content includes textual stories or educational content.

[0034] Among them, the language expression features include, but are not limited to, vocabulary selection, emotional tone, level of politeness, language style, and communication efficiency;

[0035] The behavioral characteristics include, but are not limited to, the frequency of the behavior, its persistence, selection preferences, effect evaluation, and time series.

[0036] This includes analyzing the generated language expression and behavioral characteristics data to identify the speech and behaviors that users need to improve and correct, specifically including:

[0037] Machine learning algorithms are used to learn and train on historical user data to build a model for recognizing inappropriate language and behavior.

[0038] The structured language expression features and behavioral features data are input into the trained model for pattern matching and analysis to identify users' inappropriate language and behavior.

[0039] Based on custom rules, assess whether the user's inappropriate language and behavior need improvement;

[0040] Output the analysis results, which are used to clearly identify undesirable verbal and behavioral behaviors that need to be improved.

[0041] The custom rules include frequency thresholds for the occurrence of behaviors, behavior type classifications, and behavior severity levels.

[0042] The prompts include, but are not limited to, descriptions of inappropriate language and behavior, descriptions of improvement goals, and descriptions of situational factors.

[0043] Specifically, constructing cue words based on the aforementioned undesirable speech and behaviors that need improvement includes:

[0044] Data preprocessing is performed on the undesirable speech and behaviors that need improvement;

[0045] Using a predefined library of undesirable behavior patterns, pattern matching technology is used to identify the parts of the undesirable speech and behavior that need improvement that conform to the characteristics of undesirable behavior;

[0046] Classify the identified misbehaviors;

[0047] Analyze the context in which the undesirable behavior occurs to determine the environment in which it takes place.

[0048] Improved suggestions are generated based on the classification results and context analysis results, and suggestion words are constructed.

[0049] The user behavior data refers to attendance records;

[0050] Collecting user conversation logs and user behavior data from intelligent interaction systems, specifically including:

[0051] Collecting user conversation records with intelligent interaction systems; and

[0052] By integrating with users' check-in applications, the system automatically retrieves users' check-in records.

[0053] To achieve the above objectives, the present invention also proposes a readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor of an electronic device, cause the processor to perform the steps of the intelligent interaction method described above.

[0054] To achieve the above objectives, the present invention also proposes an electronic device, comprising: a processor and a memory, wherein the memory is used to store computer program code, the computer program code including computer instructions, and when the processor executes the computer instructions, the electronic device performs the steps of the intelligent interaction method described above.

[0055] The beneficial effects of this application are as follows: Unlike existing technologies, this application provides an intelligent interactive system, method, readable storage medium, and electronic device with speech and behavior guidance functions. It extracts language expression features by analyzing user speech data and combines this with multi-dimensional analysis of behavioral data. Using custom rules, it assesses and identifies undesirable speech and behaviors that users need to improve. A prompt word construction module dynamically generates personalized prompt words, which are then used to invoke a large language model based on deep learning to generate high-quality, contextualized educational stories and teaching content. Furthermore, the system supports multimodal interaction and can provide real-time feedback on user behavior analysis results, dynamically adjust improvement strategies, and provide personalized and customized services, enhancing the user experience. This invention solves the technical problem of how to effectively monitor, analyze, and guide users to correct undesirable speech and behaviors. Through an integrated intelligent interactive system, it achieves real-time analysis and personalized feedback of user behavior, enhancing the technical effect of user experience. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0057] Figure 1 This is a schematic diagram of one embodiment of an intelligent interactive system with speech and behavior guidance function provided by the present invention;

[0058] Figure 2 This is a flowchart illustrating one embodiment of an intelligent interaction method with speech and behavior guidance function provided by the present invention;

[0059] Figure 3 yes Figure 2 A flowchart illustrating one embodiment of step S21 is shown.

[0060] Figure 4 yes Figure 2 A flowchart illustrating one embodiment of step S22 is shown;

[0061] Figure 5 yes Figure 2 A flowchart illustrating one embodiment of step S23 is shown.

[0062] Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided by the present invention.

[0063] The reference numerals used in the above figures are explained as follows:

[0064] 10. Intelligent Interaction System; 11. Data Collection Module; 12. Data Cleaning Module; 13. Data Analysis Module; 14. Prompt Word Construction Module; 15. Large Language Model Calling Module;

[0065] 3. Electronic device; 31. Processor; 32. Memory; 33. Input device; 34. Output device. Detailed Implementation

[0066] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be noted that the following embodiments are for illustrative purposes only and do not limit the scope of the application. Similarly, the following embodiments are only some, not all, embodiments of the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this application.

[0067] The terms "first," "second," and "third" in this application are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. A process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0068] In this document, the term "implementation" means that a specific feature, structure, or characteristic described in connection with an implementation may be included in at least one implementation of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same implementation, nor is it a separate or alternative implementation mutually exclusive with other implementations. It will be explicitly and implicitly understood by those skilled in the art that the implementations described herein can be combined with other implementations.

[0069] Please see Figure 1This is a schematic diagram of the structure of an intelligent interactive system with speech and behavior guidance function provided by the present invention. Specifically, the intelligent interactive system 10 includes: a data collection module 11, a data cleaning module 12, a data analysis module 13, a prompt word construction module 14, and a large language model invocation module 15. These modules work together to monitor and analyze the user's speech and behavior, promptly identify areas where the user needs improvement, and provide targeted guidance and education to the user.

[0070] The data collection module 11 is used to collect dialogue records and user behavior data between the user and the intelligent interaction system 10. The dialogue records may include various forms of data such as text and voice to accommodate different users' interaction habits and needs. The user behavior data is used to record and mark events in which the user completes specific tasks or behaviors. The collected data forms the basis for subsequent processing and analysis.

[0071] Specifically, the data collection module 11 collects dialogue records in real time by listening to the interaction interface between the user and the intelligent interaction system 10, ensuring the timeliness and continuity of the data.

[0072] In this embodiment, the user behavior data is a check-in record, which refers to the behavior recorded by a user through a check-in application after completing a specific behavior or task. The check-in record reflects the user's persistence and effort in a specific behavior, helping the system to evaluate the user's progress in improving their behavior. The check-in record includes, but is not limited to, timestamps, user identifiers, task or behavior descriptions, completion status, and user feedback or notes on the completed behavior.

[0073] The data collection module 11 integrates with the user's check-in application to automatically obtain the user's check-in records, reducing the need for manual input and improving the efficiency and accuracy of data collection.

[0074] The check-in record allows users and the system to track users' progress in developing specific behaviors or habits, such as daily exercise, study, and healthy eating. Through regular check-ins, the system can identify patterns and trends in user behavior, analyze the habit formation process based on the check-in records, and track the user's efforts and achievements in improving their behavior. The check-in records provide direct feedback on user behavior to the intelligent interaction system 10, allowing the system to offer personalized suggestions and guidance based on this data.

[0075] The collected data can be stored in a local database or a cloud server for subsequent modules to access and process.

[0076] The data collection module 11 not only collects text data but also various other forms of data, such as voice. This multimodal data collection capability allows the system to adapt to different users' interaction habits, providing a more comprehensive interactive experience. Furthermore, the data collection module 11 can collect dialogue records in real time, meaning the system can respond instantly to user input, whether text or voice. This real-time capability is key to improving user experience and system response speed. The data collection module 11 not only collects user dialogue records but also user check-in records. This comprehensive data collection can more fully reflect user behavior and habits, providing richer data support for subsequent data analysis and behavior improvement.

[0077] The data cleaning module 12 is used to analyze the dialogue records collected by the data collection module 11 to extract the user's language expression features, and to analyze the user behavior data collected by the data collection module 11 to extract the user's behavior features.

[0078] Specifically, the data cleaning module 12 analyzes the dialogue records and user behavior data collected by the data collection module 11. Through natural language processing technology and data analysis methods, it cleans and organizes the dialogue records and user behavior data collected by the data collection module 11, extracts the user's language expression features and behavioral features, and converts them into a structured data format for subsequent modules to analyze and process.

[0079] In this embodiment, the data cleaning module 12 uses NLP toolkits, such as NLTK and Spacy, to perform operations such as word segmentation, part-of-speech tagging, and named entity recognition on the dialogue records collected by the data collection module 11 in order to identify and extract language expression features; and applies data analysis methods, such as data mining techniques, to analyze user behavior data in order to identify and extract user behavior features, such as the time, location, and type of the behavior.

[0080] In this embodiment, the data cleaning module 12 converts and organizes the data format of the attendance records, extracts key information such as specific descriptions of behavior, timestamps, and user identifiers, and converts the cleaned data into structured data formats such as tables and JSON for subsequent analysis and processing by the modules.

[0081] Furthermore, the data cleaning module 12 constructs structured data from the extracted feature information, including the time, place, type (such as profanity, swearing, violent language, violent behavior, etc.) and specific process of the occurrence of inappropriate language or behavior, and transforms unstructured dialogue records and user behavior data into structured data (such as tables or JSON) so that subsequent modules can analyze and process them more effectively.

[0082] Language expression characteristics refer to a series of features exhibited by an individual in language communication, including but not limited to quantifiable language parameters such as vocabulary selection, emotional tone, politeness level, language style, and communication efficiency. In the intelligent interaction system 10, these language expression characteristics focus on the user's language usage habits in dialogue, including but not limited to:

[0083] Vocabulary selection: The range, frequency, and type of vocabulary used by the user, including technical terms, slang, profanity, etc.

[0084] Emotional expression: The emotional tendency in the user's language, such as positive, negative, angry, joyful, etc.;

[0085] Politeness level: The use of polite language in a user's speech reflects their social etiquette;

[0086] Language style: The user's language style, such as formal, informal, humorous, sarcastic, etc.;

[0087] Communication efficiency: The clarity and efficiency with which users express information;

[0088] Language errors: Errors made by users in the use of language, such as grammatical errors, inappropriate word choice, etc.

[0089] The data cleaning module 12 preprocesses the collected dialogue records, including removing noise (such as irrelevant symbols and whitespace) and standardizing the text (such as lowercase conversion and stemming). It uses NLP tools to segment the text into meaningful words or phrases. It then performs part-of-speech tagging on the segmented results to identify the part of speech (noun, verb, adjective, etc.) of each word. It identifies named entities in the text, such as personal names, place names, and organization names, which may be related to the user's emotional expression or topic. It performs sentiment analysis on the user's language expression to identify its emotional tendency, such as positive, negative, or neutral. It analyzes the user's language style to determine the formality and humor of their communication. It assesses the politeness of the user's language and identifies the use of polite language. It detects errors in the user's language use, such as grammatical errors and spelling mistakes. It extracts key language expression features from the above analysis, such as vocabulary frequency, sentiment score, and the proportion of polite language. Finally, it converts the extracted language expression features into structured data formats, such as database entries and JSON objects, for storage and further analysis.

[0090] Thus, the data cleaning module 12 can extract rich linguistic features from the dialogue records, providing accurate user language behavior data for subsequent modules of the intelligent interaction system (such as the data analysis module), thereby enabling the monitoring, analysis, and guidance of user speech and behavior. This data helps the system to understand user needs and behavioral patterns more deeply, providing users with a more personalized and effective interactive experience.

[0091] Behavioral characteristics refer to a series of quantifiable behavioral patterns exhibited by users during interactions with the intelligent interaction system, including but not limited to parameters such as frequency, persistence, preference, effect evaluation, and time series. In the intelligent interaction system 10, the behavioral characteristics focus on the user's behavioral performance during interaction with the system, including but not limited to:

[0092] Behavioral patterns: The patterns of user behavior at specific times or in specific situations, such as daily activities and habit formation;

[0093] Behavior frequency: The frequency with which a user performs a certain behavior, such as the number of times or frequency of daily check-ins;

[0094] Behavioral persistence: A user's ability to consistently perform a certain behavior over a period of time, reflecting their persistence and perseverance;

[0095] Behavioral choice: A user's preference when faced with multiple behavioral options;

[0096] Behavioral effect: The result or effect of user behavior, such as the success rate of completing a task, the effect, etc.

[0097] Behavioral time series: the chronological order in which user behaviors occur, reflecting the temporal nature of their actions.

[0098] The data cleaning module 12 preprocesses the collected user behavior data, including noise reduction, formatting, and normalization, to improve data quality; identifies user behavior events, such as check-in and task completion, and marks these events as analyzable data points; analyzes user behavior patterns, such as user behavior regularities at specific times or in specific situations; counts the frequency of a user performing a certain behavior, such as the number of times a user checks in each day; assesses a user's ability to continuously perform a certain behavior over a period of time, reflecting their persistence and perseverance; analyzes user preferences when faced with multiple behavior options; evaluates the results or effects of user behavior, such as the success rate and effectiveness of task completion; constructs a time series of user behavior to reflect the temporal nature of their behavior; extracts key behavioral features from the above analysis, such as behavior patterns, frequency, persistence, preference, effect, and time series; and converts the extracted behavioral features into structured data formats, such as database entries and JSON objects, for easy storage and further analysis.

[0099] Thus, the data cleaning module 12 can extract rich behavioral features from user behavior data, providing accurate user behavior data for subsequent modules of the intelligent interaction system (such as the data analysis module), thereby enabling the monitoring, analysis, and guidance of user behavior. This data helps the system to understand users' behavioral habits and tendencies more deeply, providing users with a more personalized and effective interactive experience.

[0100] The data analysis module 13 is used to analyze the language expression features and behavioral features data generated by the data cleaning module 12, and to identify the speech and behavior that users need to improve and correct.

[0101] Specifically, the data analysis module 13 receives language expression features and behavioral feature data processed and structured by the data cleaning module 12; it uses machine learning algorithms (such as decision trees, support vector machines, etc.) to learn and train on historical user data to establish a model for recognizing inappropriate language and behavior; it inputs the structured language expression features and behavioral feature data into the trained model for pattern matching and analysis to identify inappropriate language and behavior; it evaluates whether the user's behavior falls within the scope of needing improvement according to custom rules; and it outputs the analysis results, clearly indicating the inappropriate language and behavior that the user needs to improve, providing input for the subsequent prompt word construction module.

[0102] The data analysis module 13 uses machine learning algorithms to learn and train on historical user data to establish a model for recognizing inappropriate speech and behavior. The specific working principle includes: firstly, preprocessing the language expression features and behavioral features provided by the data cleaning module 12, which may include normalization, missing value handling, and feature selection, to ensure the data is suitable for machine learning; then training the machine learning model using historical user data, i.e., records of past user speech and behavior; this historical data has been labeled as "inappropriate" or "non-inappropriate" so that the model can learn to distinguish between these two categories. Decision trees and support vector machines (SVMs) are two commonly used machine learning algorithms that can be used to build classification models. During training, the data analysis module 13 evaluates the importance of different language expression features and behavioral features for recognizing inappropriate speech and behavior, which helps optimize model performance and may reveal which features are key indicators of inappropriate speech and behavior; the accuracy and generalization ability of the model are evaluated through cross-validation and a test set to ensure that the model performs well on unseen data; once the model training is complete, the data analysis module 13 will use these models to analyze the speech and behavior data of current users to identify inappropriate speech and behavior.

[0103] For example, the model may learn to associate certain words, phrases, or language patterns with inappropriate speech, such as profanity, insults, internet slang, and violent language. The model may also identify associations between specific behavioral patterns and inappropriate behaviors, such as stealing from others or failing to observe hygiene practices. The data analysis module 13 not only identifies inappropriate speech and behavior but also provides personalized analysis results based on the user's specific speech and behavior data, pointing out specific areas where the user needs to improve.

[0104] As new user data is continuously collected, the data analysis module 13 can continue to update and optimize the model to adapt to changes in user behavior and emerging patterns of negative speech and behavior.

[0105] The structured data processed by the data cleaning module 12, including language expression features (such as vocabulary, language style, and emotional tone) and behavioral features (such as frequency, persistence, and time series of behaviors), is input into the data analysis module 13. The data analysis module 13 uses pre-trained machine learning models (such as decision trees and support vector machines) trained on historical user data to identify patterns of inappropriate speech and behavior. It matches the current user's structured speech and behavior data with the inappropriate patterns learned in the model. This process involves comparing the features of the user data with the features in the model to determine if a match exists. Through pattern matching, the data analysis module 13 analyzes the user's speech and behavior data, identifying data points that match inappropriate patterns. For example, if a user's speech contains profanity or insults, or if the user's behavior data indicates theft, these will be marked as inappropriate. The data analysis module 13 outputs analysis results, clearly indicating which of the user's speech and behaviors are inappropriate. These results typically include detailed information such as the type of inappropriate speech and behavior, the time of occurrence, and the frequency.

[0106] The custom rules may include frequency thresholds for the behavior, behavior type classifications, and behavior severity levels. The data analysis module 13 applies these custom rules, comparing user behavior data with the parameters and conditions defined in the custom rules to determine whether the user behavior falls within the scope requiring improvement. The custom rules can be dynamically adjusted based on new data input and system feedback to optimize the performance of the evaluation mechanism.

[0107] For example, the data analysis module 13 analyzes the frequency of undesirable behaviors and determines whether the frequency exceeds a threshold to identify frequently made mistakes that require priority improvement. The data analysis module 13 assesses the severity of undesirable behaviors, identifying which behaviors have a greater impact on the user or others and require more urgent attention. Based on built-in rules or algorithms, the data analysis module 13 determines which behaviors need immediate improvement and which can be addressed later.

[0108] Suppose our intelligent interactive system aims to help users improve their personal health habits, especially focusing on users' bedtime hygiene habits, such as brushing their teeth before bed.

[0109] Custom rule settings are as follows:

[0110] Frequency threshold for the behavior: The rule requires users to brush their teeth at least 6 nights a week. If they brush their teeth less than this number, it is considered a bad behavior.

[0111] Behavioral type classification: "Brushing teeth before bed" is classified as a "healthy habit".

[0112] Severity level of behavior: If a user does not brush their teeth for three consecutive days, the system will mark it as a "high severity" behavior.

[0113] The system collects users' pre-sleep brushing data through an integrated toothbrush sensor or user check-in records in a mobile application. The system monitors users' brushing behavior every night, recording whether brushing occurred and the time. The data analysis module 13 evaluates the user's brushing behavior according to custom rules. For example, if the system finds that a user only brushes their teeth four nights a week, which is below the set threshold of six days, the system identifies this behavior as matching a preset pattern of poor behavior and therefore marks it as a behavior that needs improvement. If a user does not brush their teeth for two consecutive days, the system further assesses the severity and marks it as a "high-severity" behavior if brushing is still not performed on the third day.

[0114] Based on user feedback and data on behavioral changes, the system may find that the current 6-day threshold is too strict for some users. Therefore, the system can adjust the threshold according to the actual situation, for example, to 5 days, to more reasonably accommodate the needs of different users.

[0115] The prompt word construction module 14 is used to construct prompt words based on the undesirable speech and behavior that need to be improved identified by the data analysis module 13; wherein, the prompt words include, but are not limited to, descriptions of undesirable speech and behavior, descriptions of improvement goals, and descriptions of situational factors.

[0116] Specifically, the prompts include detailed descriptions of users' inappropriate language and behavior, such as "using vulgar language in public" or "stealing other people's belongings in a shopping mall." The prompts also include descriptions of improvement goals, i.e., the specific actions or states the user should take, such as "using polite language" or "respecting other people's property." Furthermore, the prompts consider descriptions of the contextual factors that cause the inappropriate behavior, ensuring that the subsequently generated stories or suggestions are relevant to the user's actual experience, such as "being patient while shopping in a mall."

[0117] The prompt word construction module 14 combines the user's personal characteristics and behavioral history, and generates personalized improvement suggestions based on the user's negative speech and behavior using natural language generation technology (such as Transformer, GPT-2, etc.). These suggestions aim to help the user understand why improvement is needed and how to improve.

[0118] The prompt word construction module 14 uses a predefined library of bad behavior patterns to identify the parts of user behavior that conform to the characteristics of bad behavior through pattern matching technology; it classifies the identified bad behaviors, such as inappropriate language and inappropriate behavior in public places, so as to generate targeted improvement prompts; it also analyzes the context in which bad behaviors occur to understand in what environment these behaviors usually occur, so as to generate more specific improvement prompts.

[0119] Furthermore, the prompts also include a behavior correction guidance script; wherein, the behavior correction guidance script includes a series of preset plots, characters, and dialogues, which can be adjusted according to the user's specific situation to ensure the relevance and appeal of the story. The behavior correction guidance script is a technical implementation designed to provide guidance and demonstrations for behavior correction by simulating real or fictional scenarios; through narrative, it immerses the user in a specific environment, allowing the user to emotionally and cognitively resonate with the characters in the story, thereby more effectively learning and correcting undesirable behaviors.

[0120] For users' speech and behaviors that need improvement and correction, a story is explicitly generated within the prompts to teach users how to correct inappropriate speech and behavior and to provide examples of correct speech and behavior. The prompts can be constructed by combining the user's specific situation and needs, using natural language generation technology to generate targeted and attractive prompts. The stories can be presented in a vivid and interesting way, making them easier for users to accept and understand, thereby motivating users to correct their undesirable behavior.

[0121] The large language model invocation module 15 is used to invoke the large language model based on the prompt words constructed by the prompt word construction module 14 to generate corresponding contextualized content. The contextualized content includes text stories or educational content; depending on the user's needs, the generated content can be in text form or further converted into speech form to adapt to different usage scenarios.

[0122] Contextualized content generation refers to the process of generating customized text content based on specific contextual information, such as a user's inappropriate behavior and speech, using a large language model. This content aims to provide guidance for education and behavior correction.

[0123] Specifically, the large language model invocation module 15 receives prompts generated by the prompt word construction module 14. These prompts contain information such as descriptions of the user's inappropriate behavior, improvement goals, and story themes. Using an API interface or by directly calling the large language model's code library, the prompts are used as input to request the large language model to generate corresponding stories or educational content. Based on the input prompts, the large language model utilizes its deep learning algorithms and text generation capabilities trained on large-scale language data to generate text stories or educational content that match the context of the prompts.

[0124] For example, when the large language model invocation module 15 receives prompt words such as "public place," "control emotions," or "polite language," it invokes a large language model via an API interface, inputting the prompt words to request story generation. The large language model generates a story about a person who swears in public and eventually learns to control their emotions. Depending on user preferences, the story can be displayed as text on the user's device or converted to speech and played through a smart speaker.

[0125] Suppose that the intelligent interaction system 10 detects that a user is using foul language in a shopping mall, it needs to construct a prompt word and generate a story to teach the user not to use foul language and improve their behavior in public.

[0126] The data analysis module 13 identifies the user's use of profanity in the shopping mall and marks it as a negative behavior requiring improvement. The prompt word construction module 14 determines improvement goals based on this behavior, namely, reducing or eliminating the use of profanity in public. The prompt word construction module 14 constructs prompt words that clearly indicate the negative behavior (using profanity) and the improvement goal (controlling emotions, using polite language), and accordingly constructs a behavior correction guidance script: The story tells of an anthropomorphic animal character who, while playing in a park (a location the user prefers), uses impolite language due to a minor disagreement, subsequently realizes their mistake, and corrects it. Based on the user's age and gender, the system selects the most appropriate voice and visual presentation. Assuming the user is a young woman, the system selects a deep, resonant male voice to narrate the story and displays illustrations on devices with screens. The story is presented to the user in audio and visual form via smart speakers or mobile applications; the audio is a deep, resonant male voice, and the screen displays illustrations matching the storyline, enhancing the story's appeal and educational effect.

[0127] As described above, this application provides an intelligent interactive system with speech and behavior guidance capabilities. It extracts language expression features by analyzing user speech data and combines this with multi-dimensional analysis of behavioral data. Using custom rules, it assesses and identifies inappropriate speech and behaviors that users need to improve. A prompt word construction module dynamically generates personalized prompt words, which are then used to invoke a deep learning-based large language model to generate high-quality, contextualized educational stories and teaching content. Furthermore, the system supports multimodal interaction and can provide real-time feedback on user behavior analysis results, dynamically adjust improvement strategies, and provide personalized and customized services, enhancing the user experience. This invention solves the technical problem of how to effectively monitor, analyze, and guide users to correct inappropriate speech and behaviors. Through an integrated intelligent interactive system, it achieves real-time analysis and personalized feedback of user behavior, enhancing the technical effect of user experience.

[0128] Please see Figure 2 This is a flowchart illustrating the first embodiment of an intelligent interaction method with speech and behavior guidance function provided by the present invention. The method includes the following steps:

[0129] Step S20: Collect dialogue records and user behavior data between the user and the intelligent interaction system 10.

[0130] The dialogue records can include various forms of data such as text and voice to accommodate different users' interaction habits and needs. User behavior data is used to record and mark events in which users complete specific tasks or behaviors. The collected data forms the basis for subsequent processing and analysis.

[0131] Specifically, by monitoring the interaction interface between the user and the intelligent interaction system 10, dialogue records are collected in real time to ensure the timeliness and continuity of the data; and by integrating with the user's check-in application to obtain the user's check-in records, the need for manual input is reduced, and the efficiency and accuracy of data collection are improved.

[0132] In this embodiment, the user behavior data is a check-in record, which refers to the behavior recorded by a user through a check-in application after completing a specific behavior or task. The check-in record reflects the user's persistence and effort in a specific behavior, helping the system to evaluate the user's progress in improving their behavior. The check-in record includes, but is not limited to, timestamps, user identifiers, task or behavior descriptions, completion status, and user feedback or notes on the completed behavior.

[0133] The check-in record allows users and the system to track users' progress in developing specific behaviors or habits, such as daily exercise, study, and healthy eating. Through regular check-ins, the system can identify patterns and trends in user behavior, analyze the habit formation process based on the check-in records, and track the user's efforts and achievements in improving their behavior. The check-in records provide direct feedback on user behavior to the intelligent interaction system 10, allowing the system to offer personalized suggestions and guidance based on this data.

[0134] Step S21: Analyze the collected dialogue records to extract the user's language expression features, and analyze the collected user behavior data to extract the user's behavioral features.

[0135] Please also refer to Figure 3 Step S21, namely, analyzing the collected dialogue records to extract the user's language expression features and analyzing the collected user behavior data to extract the user's behavioral features, specifically includes:

[0136] Step S210: Preprocess the collected dialogue records;

[0137] Specifically, preprocessing includes noise removal (such as irrelevant symbols, whitespace, etc.) and text standardization (such as uniform lowercase conversion, stemming, etc.).

[0138] Step S211: Process the collected dialogue records to identify and extract language expression features;

[0139] The language expression features include, but are not limited to, quantifiable language parameters such as vocabulary selection, emotional tone, politeness level, language style, and communication efficiency.

[0140] In this embodiment, Natural Language Processing (NLP) techniques (such as NLTK and Spacy) are used to perform word segmentation, part-of-speech tagging, and named entity recognition on the collected dialogue records to identify and extract language expression features. Data analysis methods (such as data mining techniques) are used to analyze user behavior data to identify and extract user behavioral characteristics.

[0141] Specifically, NLP tools are used to segment the text, breaking down continuous text into meaningful words or phrases; the segmented results are then labeled with part-of-speech tags to identify the part of speech (noun, verb, adjective, etc.) of each word; named entities in the text, such as names of people, places, and organizations, are identified; sentiment analysis is performed on the user's language expression to identify their emotional tendencies; the user's language style is analyzed to determine the formality and humor of their communication; the politeness level in the user's language is assessed to identify the use of polite language; and errors in the user's language use, such as grammatical errors and spelling mistakes, are detected.

[0142] Key language expression features are extracted from the above analysis, such as vocabulary usage frequency, sentiment score, and proportion of polite language. The extracted language expression features are then transformed into structured data formats, such as database entries and JSON objects, for storage and further analysis.

[0143] Step S212: Process the collected user behavior data to extract user behavior characteristics.

[0144] The behavioral characteristics are quantifiable behavioral patterns exhibited by the user during interaction with the intelligent interaction system 10, including but not limited to parameters such as the frequency of behavior, duration, selection preferences, effect evaluation, and time series.

[0145] In this embodiment, data analysis methods (such as data mining techniques) are used to analyze user behavior data in order to identify and extract user behavioral characteristics.

[0146] In this embodiment, the user behavior data is the open record. The collected check-in records are converted and organized in data format, and key information such as the specific description of the behavior, timestamp, user identifier, etc. are extracted to obtain the user's behavioral characteristics.

[0147] Specifically, the process involves identifying user behavior events, such as checking in and completing tasks, and marking these events as analyzable data points; analyzing patterns in user behavior, such as behavioral patterns in specific times or contexts; counting the frequency of a user performing a certain behavior, such as the number of times a user checks in each day; assessing a user's ability to consistently perform a certain behavior over a period of time; analyzing a user's preference when faced with multiple behavioral options; evaluating the results or effects of user behavior, such as the success rate and effectiveness of completing tasks; and constructing a time series of user behavior to reflect the temporal nature of their actions.

[0148] Key behavioral features, such as behavioral patterns, frequency, persistence, choice preferences, effects, and time series, are extracted from the above analysis. These features are then transformed into structured data formats, such as database entries and JSON objects, for storage and further analysis.

[0149] Step S213: Construct structured data from the extracted language expression features and behavioral features, and construct structured data from the collected dialogue records and user behavior data.

[0150] Step S22: Analyze the generated language expression features and behavioral features data to identify the speech and behavior that the user needs to improve and correct.

[0151] Please also refer to Figure 4 Step S22 involves analyzing the generated language expression and behavioral feature data to identify the user's speech and behaviors that need improvement and correction, specifically including:

[0152] Step S220: Use machine learning algorithms to learn and train on historical user data to establish a model for recognizing inappropriate language and behavior.

[0153] The machine learning algorithms mentioned include decision trees and support vector machines.

[0154] Specifically, historical user data—records of past user speech and behavior—is used to train a machine learning model. This historical data is labeled as "inappropriate" or "non-inappropriate" so the model can learn to distinguish between these two categories. During model training, the importance of different language expression features and behavioral features in identifying inappropriate speech and behavior is evaluated. The model's accuracy and generalization ability are assessed through cross-validation and a test set to ensure good performance on unseen data. Once training is complete, these models will be used to analyze current user speech and behavior data to identify inappropriate speech and behavior.

[0155] Step S221: Input the structured language expression features and behavioral features data into the trained model for pattern matching and analysis to identify the user's inappropriate speech and behavior.

[0156] Specifically, using a pre-trained machine learning model, the structured language expression and behavioral features of the current user are matched against maladaptive patterns learned in the model. This process involves comparing the features of the user data with those in the model to determine if a match exists. Through pattern matching, data points that match maladaptive patterns are identified. The analysis results are output, clearly indicating the user's maladaptive speech and behavior. These results typically include detailed information such as the type, timing, and frequency of the maladaptive speech and behavior.

[0157] Step S222: Based on the custom rules, assess whether the user's inappropriate language and behavior need improvement. If yes, proceed to step S223; otherwise, return to step S220.

[0158] The custom rules may include frequency thresholds for the occurrence of the behavior, behavior type classifications, and behavior severity levels.

[0159] Specifically, by comparing the user's inappropriate language and behavior with the parameters and conditions defined in the custom rules, it is determined whether the user's inappropriate language and behavior needs improvement. The custom rules can be dynamically adjusted based on new data input and system feedback to optimize the performance of the evaluation mechanism.

[0160] The custom rules contain parameters and conditions for inappropriate speech and behavior. These rules are implemented in code and stored in a rule base. By collecting user speech and behavior data, rules from the rule base are loaded and matched against preprocessed user data. For each rule, the system evaluates whether the user's speech and behavior meet the conditions defined in the rule. For example, if a rule defines "using profanity in public" as inappropriate behavior, the system checks if such behavior exists in the user data. Based on the results of the condition evaluation, the system determines whether the user's speech and behavior need improvement. If a user's behavior triggers a condition in the rule, the system marks that behavior as needing improvement.

[0161] Step S223: Output the analysis results, which are used to clearly identify the undesirable speech and behavior that need to be improved.

[0162] Step S23: Construct cue words based on the negative speech and behavior that need to be improved.

[0163] The prompts include, but are not limited to, descriptions of inappropriate language and behavior, descriptions of improvement goals, descriptions of situational factors, and behavior correction guidance scripts.

[0164] Specifically, the prompts include detailed descriptions of inappropriate language and behavior, such as "using vulgar language in public" or "stealing other people's belongings in a shopping mall." The prompts also include descriptions of improvement goals, i.e., the specific actions or states the user should take, such as "using polite language" or "respecting other people's property." Furthermore, the prompts consider descriptions of the contextual factors that cause the inappropriate behavior, ensuring that the subsequently generated stories or suggestions are relevant to the user's actual experience, such as "being patient while shopping in a mall."

[0165] The behavior correction guidance script includes a series of pre-set plots, characters, and dialogues. These elements can be adjusted according to the user's specific situation to ensure the relevance and appeal of the story. The behavior correction guidance script is a technological implementation designed to provide guidance and demonstrations for behavior correction by simulating real or fictional scenarios. Through narrative, it immerses the user in a specific environment, allowing the user to emotionally and cognitively resonate with the characters in the story, thereby more effectively learning and correcting undesirable behaviors.

[0166] Please also refer to Figure 5 Step S23, constructing cue words based on the unpleasant speech and behavior that need improvement, specifically including:

[0167] Step S230: Perform data preprocessing on the undesirable speech and behavior that need improvement.

[0168] Specifically, based on the problematic speech and behavior that need improvement, this data contains detailed analysis results of user speech and behavior, identifying which speech and behavior are identified as problematic; the data is preprocessed, including data cleaning (removing noise and irrelevant information), formatting (unifying the data format), and normalization (making the data comparable on the same scale).

[0169] Step S231: Using a predefined library of undesirable behavior patterns, identify the parts of the undesirable speech and behavior that need improvement that conform to the characteristics of undesirable behavior through pattern matching technology.

[0170] Step S232: Classify the identified undesirable behaviors.

[0171] For example, inappropriate language, inappropriate behavior in public places, etc., so as to generate targeted improvement suggestions;

[0172] Step S233: Analyze the context in which the undesirable behavior occurs to determine the environment in which the undesirable behavior occurs.

[0173] Step S234: Generate improved prompts based on the classification results and context analysis results, and construct prompt words.

[0174] Furthermore, by combining users' personal characteristics and behavioral history, and based on users' negative language and behavior, natural language generation technologies (such as Transformer, GPT-2, etc.) are used to generate personalized improvement suggestions. These suggestions are designed to help users understand why improvements are needed and how to improve.

[0175] Step S24: Based on the constructed prompt words, call the large language model to generate corresponding contextualized content.

[0176] The contextualized content includes text stories or educational content; the contextualized content generated according to the user's needs can be in text form or further converted into audio form to adapt to different usage scenarios.

[0177] Contextualized content generation refers to the process of generating customized text content based on specific contextual information, such as a user's inappropriate behavior and speech, using a large language model. This content aims to provide guidance for education and behavior correction.

[0178] Specifically, using an API interface or directly calling the code library of a large language model, the system takes the prompt words as input and requests the large language model to generate corresponding correction guidance scripts. Based on the input prompt words, the large language model leverages its deep learning algorithms and text generation capabilities trained on large-scale language data to generate text stories or educational content that match the context of the prompt words. The generated content is tailored to the user's specific situation, ensuring the relevance and effectiveness of the story or educational content. The final output text story or educational content directly addresses the user's undesirable behaviors and speech, providing specific and personalized correction suggestions and examples.

[0179] In this embodiment, the advantages and beneficial effects of the intelligent interaction method with speech and behavior guidance function have been described above and will not be repeated here. Since the intelligent interaction method with speech and behavior guidance function is applied to the intelligent interaction system with speech and behavior guidance function, the intelligent interaction method with speech and behavior guidance function also has the same advantages and beneficial effects.

[0180] One embodiment of the present invention also provides a readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor of an electronic device, cause the processor to perform the steps of any of the above-described intelligent interaction methods with speech guidance functions.

[0181] One embodiment of the present invention also provides an electronic device, including: a processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, and when the processor executes the computer instructions, the electronic device performs the steps of any of the above-described intelligent interaction methods with speech guidance function.

[0182] Please see Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention.

[0183] The electronic device 3 includes a processor 31, a memory 32, an input device 33, and an output device 34. The processor 31, memory 32, input device 33, and output device 34 are coupled together via connectors, which may include various interfaces, transmission lines, or buses, etc., and are not limited in this respect in the embodiments of the present invention. It should be understood that in the various embodiments of the present invention, coupling refers to mutual connection through a specific method, including direct connection or indirect connection through other devices, such as through various interfaces, transmission lines, buses, etc.

[0184] The processor 31 can be one or more graphics processing units (GPUs). If the processor 31 is a GPU, the GPU can be a single-core GPU or a multi-core GPU. Optionally, the processor 31 can be a processor group composed of multiple GPUs, with the multiple processors coupled to each other via one or more buses. Optionally, the processor can also be other types of processors, etc., and the embodiments of the present invention are not limited thereto.

[0185] The memory 32 can be used to store computer program 35, as well as various types of computer program code, including program code for executing the present invention. Optionally, the memory may include at least random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM), which is used for related instructions and data.

[0186] Input device 33 is used to input data and / or signals, and output device 34 is used to output data and / or signals. Output device 33 and input device 34 can be independent devices or an integrated device.

[0187] It is understood that in this embodiment of the invention, the memory 32 can be used not only to store related instructions, but also the specific data stored in the memory is not limited.

[0188] Understandable, Figure 6This is merely a simplified design of an electronic device. In practical applications, the electronic device may also include other necessary components, including, but not limited to, any number of input / output devices, processors, memories, etc., and all control units of underwater robots that can implement embodiments of the present invention are within the scope of protection of the present invention.

[0189] The above description is only a partial embodiment of this application and does not limit the scope of protection of this application. Any equivalent device or equivalent process transformation made based on the content of this application specification and drawings, or directly or indirectly applied to other related technical fields, are similarly included within the scope of patent protection of this application.

Claims

1. An intelligent interactive system, characterized in that, include: The data collection module is used to collect dialogue records between users and the intelligent interaction system, as well as user behavior data; among which, user behavior data is used to record and mark events in which users complete specific tasks or behaviors. The data cleaning module is used to analyze the dialogue records collected by the data collection module to extract the user's language expression features, and to analyze the user behavior data collected by the data collection module to extract the user's behavior features. The data analysis module is used to analyze the language expression features and behavioral features data generated by the data cleaning module, and to identify the speech and behavior of users that need to be improved and corrected. The prompt word construction module is used to construct prompt words based on the undesirable speech and behaviors requiring improvement identified by the data analysis module; wherein, the prompt words include a behavior correction guidance script, and the behavior correction guidance script includes preset scenarios, roles, and dialogues; and The large language model invocation module is used to invoke the large language model based on the prompt words constructed by the prompt word construction module to generate corresponding contextualized content; wherein, the contextualized content includes text stories or educational content.

2. The intelligent interactive system according to claim 1, characterized in that, The language expression features include, but are not limited to, vocabulary selection, emotional tone, level of politeness, language style, and communication efficiency; The behavioral characteristics include, but are not limited to, the frequency of the behavior, its persistence, selection preferences, effect evaluation, and time series.

3. The intelligent interactive system according to claim 2, characterized in that, The data analysis module is used for: Receive language expression features and behavioral features data processed and structured by the data cleaning module; Machine learning algorithms are used to learn and train on historical user data to build a model for recognizing inappropriate language and behavior. The structured language expression features and behavioral features data are input into the trained model for pattern matching and analysis to identify inappropriate speech and behavior. Based on custom rules, assess whether user behavior needs improvement; The output analysis results clearly indicate the user's inappropriate language and behavior that needs improvement.

4. The intelligent interactive system according to claim 3, characterized in that, The custom rules include frequency thresholds for the occurrence of the behavior, behavior type classifications, and behavior severity levels.

5. The intelligent interactive system according to claim 1, characterized in that, The prompts include, but are not limited to, descriptions of inappropriate language and behavior, descriptions of improvement goals, and descriptions of situational factors.

6. The intelligent interactive system according to claim 5, characterized in that, The prompt word construction module is used for: Data preprocessing is performed on the undesirable speech and behaviors that need improvement; Using a predefined library of undesirable behavior patterns, pattern matching technology is used to identify the parts of the undesirable speech and behavior that need improvement that conform to the characteristics of undesirable behavior; Classify the identified misbehaviors; Analyze the context in which the undesirable behavior occurs to determine the environment in which it takes place. Improved suggestions are generated based on the classification results and context analysis results, and suggestion words are constructed.

7. The intelligent interactive system according to claim 1, characterized in that, The user behavior data is the attendance record; The data collection module is used to automatically obtain the user's check-in records by integrating with the user's check-in application.

8. An intelligent interaction method, characterized in that, The method includes: Collect dialogue records between users and intelligent interaction systems, as well as user behavior data; wherein, the user behavior data is used to record and mark events in which users complete specific tasks or behaviors; Analyze collected dialogue records to extract users' language expression features, and analyze collected user behavior data to extract users' behavioral features; Analyze the generated language expression and behavioral feature data to identify the speech and behavior of users that need improvement and correction; Based on the identified inappropriate speech and behavior requiring improvement, cue words are constructed; wherein, the cue words include a behavior correction guidance script, and the behavior correction guidance script includes preset scenarios, roles, and dialogues; The large language model is invoked based on the constructed prompt words to generate corresponding contextualized content; wherein, the contextualized content includes textual stories or educational content.

9. The intelligent interaction method according to claim 8, characterized in that, The language expression features include, but are not limited to, vocabulary selection, emotional tone, level of politeness, language style, and communication efficiency; The behavioral characteristics include, but are not limited to, the frequency of the behavior, its persistence, selection preferences, effect evaluation, and time series.

10. The intelligent interaction method according to claim 9, characterized in that, Analyze the generated language expression and behavioral feature data to identify the speech and behaviors that users need to improve and correct, specifically including: Machine learning algorithms are used to learn and train on historical user data to build a model for recognizing inappropriate language and behavior. The structured language expression features and behavioral features data are input into the trained model for pattern matching and analysis to identify users' inappropriate language and behavior. Based on custom rules, assess whether the user's inappropriate language and behavior need improvement; Output the analysis results, which are used to clearly identify undesirable verbal and behavioral behaviors that need to be improved.

11. The intelligent interaction method according to claim 10, characterized in that, The custom rules include frequency thresholds for the occurrence of the behavior, behavior type classifications, and behavior severity levels.

12. The intelligent interaction method according to claim 8, characterized in that, The prompts include, but are not limited to, descriptions of inappropriate language and behavior, descriptions of improvement goals, and descriptions of situational factors.

13. The intelligent interaction method according to claim 12, characterized in that, Based on the aforementioned undesirable speech and behaviors that need improvement, cue words are constructed, specifically including: Data preprocessing is performed on the undesirable speech and behaviors that need improvement; Using a predefined library of undesirable behavior patterns, pattern matching technology is used to identify the parts of the undesirable speech and behavior that need improvement that conform to the characteristics of undesirable behavior; Classify the identified misbehaviors; Analyze the context in which the undesirable behavior occurs to determine the environment in which it takes place. Improved suggestions are generated based on the classification results and context analysis results, and suggestion words are constructed.

14. The intelligent interactive system according to claim 8, characterized in that, The user behavior data is the attendance record; Collecting user conversation logs and user behavior data from intelligent interaction systems, specifically including: Collect user conversation records with intelligent interaction systems; as well as By integrating with users' check-in applications, the system automatically retrieves users' check-in records.

15. A readable storage medium storing a computer program, characterized in that, The computer program includes program instructions that, when executed by the processor of the electronic device, cause the processor to perform the steps of the intelligent interaction method as described in any one of claims 8 to 14.

16. An electronic device comprising: A processor and a memory, characterized in that the memory is used to store computer program code, the computer program code including computer instructions, wherein when the processor executes the computer instructions, the electronic device performs the steps of the intelligent interaction method as described in any one of claims 8 to 14.

Images