Chat system with functions of user psychological exclusive virtual interaction and emotional value provision

Abstract

The invention discloses a chat system with functions of user psychological exclusive virtual interaction and emotional value providing, and belongs to the field of virtual interaction. Comprising a registration and login module, a psychological analysis module, a personality generation module, an interaction matching module, an interaction cultivation module, an exclusive management module, a stage management module, a relation authentication module, an account logout module and an offline interaction module. According to the method, the interaction between the user and the most matched AI robot can be ensured, the real and stable social experience can be provided, the chapter interaction feeling is reduced, the user belonging feeling and the interaction viscidity are enhanced, and separation and new connection in real social contact can be simulated more truly.

Description

Technical Field

[0001] The present invention relates to the field of virtual interaction, and particularly to a chat system with user psychological exclusivity virtual interaction and emotional value provision. Background Art

[0002] In recent years, with the progress of artificial intelligence (AI) technology, the way of human-computer interaction has undergone a major transformation. Especially today, with the highly developed informatization and intelligence, the relationship between humans and machines is becoming closer. Traditional AI chat systems mainly focus on information retrieval, task execution, and daily communication, but lack deep emotional connection and personalized experience, and it is difficult to meet the psychological needs of users. For lonely individuals, people with greater psychological pressure, or users who hope to obtain a more companionable interaction experience, existing AI systems cannot meet their deeper emotional needs. Therefore, a chat system with user psychological exclusivity virtual interaction and emotional value provision has emerged as the times require.

[0003] Existing chat systems with user psychological exclusivity virtual interaction and emotional value provision cannot provide a real and stable social experience, increase the sense of stereotyped interaction, reduce user belonging and interaction stickiness, and cannot truly simulate separation and new connections in real social interactions. For this reason, we propose a chat system with user psychological exclusivity virtual interaction and emotional value provision. Summary of the Invention

[0004] The purpose of the present invention is to solve the defects existing in the prior art, and to propose a chat system with user psychological exclusivity virtual interaction and emotional value provision.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A chat system with user psychological exclusivity virtual interaction and emotional value provision, including a registration and login module, a psychological analysis module, a personality generation module, an interaction matching module, an interaction cultivation module, an exclusivity management module, a stage management module, a relationship authentication module, an account cancellation module, and an offline interaction module;

[0007] The registration and login module is used for users to register and upload personal information, and to log in to the chat system;

[0008] The psychological analysis module is used to comprehensively evaluate the user's personality characteristics, psychological needs, and communication preferences, and to construct a corresponding user portrait;

[0009] The personality generation module is used to construct a unique growth experience for each AI robot and to train each AI robot;

[0010] The interaction matching module is used to recommend AI robots with a matching style for users according to the user evaluation results;

[0011] The interactive cultivation module supports the user to have real-time text, voice or video chats with the AI robot and continuously optimizes the dialogue content of the AI robot;

[0012] The exclusive management module is used to evaluate the intimacy of the chat between the AI robot and the corresponding user and determine whether the user can exclusively occupy the AI for a long time;

[0013] The stage management module is used to analyze the chat data between the user and the corresponding AI robot and adjust the interaction strategy of the AI;

[0014] The relationship authentication module is used to conduct virtual partner authentication for users and AI robots with high intimacy;

[0015] The account cancellation module is used for the user to apply for permanent cancellation of any AI robot;

[0016] The offline interaction module is used to assign a real person to interact with the user offline based on the chat content between the user and the AI robot.

[0017] As a further solution of the present invention, the specific steps for the psychological analysis module to construct the corresponding user portrait are as follows:

[0018] S1.1: The psychological analysis module collects various user data, including age, gender, occupation, educational background, language style, psychological characteristics, communication preferences, and behavior data, and establishes a corresponding user feature set based on the collected user data. Then, using the Big Five personality model, the user's personality is divided into five dimensions: openness O, conscientiousness C, extraversion E, agreeableness A, and neuroticism N;

[0019] S1.2: Collect the registered user feature sets of each user, extract the corresponding groups of personality feature data points from each user feature set to construct a data matrix, calculate the principal component vectors and their weights through PCA, extract the proportion of the five major personality principal components, and generate the user's personality characteristics;

[0020] S1.3: Use NLP emotion analysis to divide the user's historical chat record text into positive words, negative words, and neutral words, and then calculate the proportion of each type of word in the total number of words to obtain the user's positive emotion ratio, negative emotion ratio, and neutral emotion ratio. Then, combine with neuroticism N to calculate the user's average emotional fluctuation;

[0021] S1.4: Count the number of occurrences of each word in the user's chat records in the total chat record text, calculate its corresponding word frequency, then calculate the inverse document frequency of each word through the TF-IDF model, identify the topics that the user most often talks about based on the output results of the TF-IDF model, and calculate the importance of each topic to generate the user's topic interest;

[0022] S1.5: Construct a complete user profile based on the generated user feature set, user personality traits, average user mood swings, and user topic interests. Then, regularly collect the chat records between the user and the AI robot and update the user profile in real time.

[0023] As a further solution of the present invention, the specific steps for the personality generation module to train each AI robot are as follows:

[0024] S2.1: The personality generation module assigns a unique identification ID to each AI robot. When each AI robot is born, randomly initialize the growth environment and innate cognitive level of each AI robot, and initialize the initial personality of the AI robot based on the set growth environment template;

[0025] S2.2: Conduct growth simulations on the AI robots, and based on the change of the time in the system, dynamically adjust the growth environment and innate cognitive level of each AI robot in a randomly selected manner, and generate unique growth experiences for each AI robot;

[0026] S2.3: Adjust the personalities of each AI robot according to the changes in each growth environment, innate cognitive level, and its own experiences, and form a specific language style, including sense of humor, emotional expression methods, formal expression, and informal expression;

[0027] S2.4: Constrain the knowledge ability of the AI according to the growth experiences in the growth simulations of each AI robot, and construct a corresponding knowledge shielding matrix to control the accessible knowledge range of the AI. Then, when the growth simulations of each AI robot reach their respective preset simulation times, stop the simulations. The personality generation module freezes the personalities of each AI robot and generates corresponding AI robot portraits, and at the same time, updates the AI robot portraits in real time according to the subsequent chat records between each AI robot and the user.

[0028] As a further solution of the present invention, the specific steps for the interaction matching module to recommend an AI robot with a matching style for the user are as follows:

[0029] S3.1: Before the matching starts, calculate the cosine similarity between each user profile and the AI robot portrait, use the similarity between the user and the AI robot as its pheromone concentration, and construct an initial pheromone matrix. Initialize a group of populations and set the number of explorers in the population, where each explorer represents a matching task;

[0030] S3.2: Treat each user portrait as the starting point, initialize the starting position of the exploration body. After that, each exploration body calculates the selection probability of each AI robot portrait based on the user portrait corresponding to the current starting point, and normalizes the selection probabilities of all candidate AI robot portraits. Then, generate a random number and calculate the cumulative probability distribution, and select the AI robot portrait through the roulette method;

[0031] S3.3: After the matching is completed, calculate its fitness value according to the similarity between the selected AI robot portrait and the user portrait, and update the pheromone concentration according to the fitness value. Then, repeat multiple rounds of selection and pheromone concentration update until the change value of the pheromone concentration converges within the preset threshold range after multiple iterations, and then stop the iteration. In the final round, select the matching pair with the highest pheromone concentration as the recommendation result;

[0032] S3.4: Mark the detailed information of the corresponding AI robot portrait in the matching pair as the recommended chat object, and feedback the marked AI robot and the unmarked AI robots to the user interface for the user to select.

[0033] As a further solution of the present invention, the specific calculation formula of the cosine similarity described in S3.1 is as follows:

[0034]

[0035] In the formula, S ij represents the similarity between the i-th user and the j-th AI robot, and the larger the similarity value, the higher the matching degree; U i represents the user portrait vector of the i-th user; A j represents the robot portrait vector of the j-th AI robot; ||U i || represents the modulus length of the user portrait; ||A j || represents the modulus length of the AI robot portrait;

[0036] The specific calculation formula of the selection probability described in S3.2 is as follows:

[0037] η ij =S ij

[0038]

[0039] In the formula, η ij represents the heuristic information; P ij represents the probability of selecting the j-th AI robot from the i-th user; p ij represents the current pheromone concentration; α and β respectively represent the weights of controlling the pheromone and the heuristic information; k represents the index of each AI robot;

[0040] The specific calculation formula for updating the pheromone concentration described in S3.3 is as follows:

[0041]

[0042] In the formula, f(U i , A j ) represents the fitness function; Q represents the pheromone enhancement factor; represents the cumulative pheromone contributed by all exploration bodies; ρ represents the pheromone evaporation coefficient; m represents the total number of exploration bodies.

[0043] As a further solution of the present invention, the specific steps for the interactive training module to optimize the conversation content of the AI robot are as follows:

[0044] S4.1: Real-time statistics of the number of messages sent by the user to the selected AI robot within a preset time interval, and calculation of the chat frequency between the user and the AI robot in each time period. If the user's chat frequency is higher than the preset threshold, it indicates that the user is highly active and enters the close communication stage. Otherwise, it indicates a decrease in interaction and enters the cold period;

[0045] S4.2: Collect the chat records between the user and the corresponding AI robot, identify the emotional words in each message in the chat records through NLP technology, and calculate the emotional polarity score of the corresponding message text. Then, based on the emotional intensity weight of the message, obtain the user's emotional investment degree in this chat. If the emotional investment degree is higher than the preset threshold, indicating high emotional investment, it indicates entering the intimate stage. Otherwise, it indicates entering the cold stage;

[0046] S4.3: Calculate the interaction content matching degree of each conversation between the user and the AI robot through the cosine similarity formula. If the interaction content matching degree is higher than the preset threshold, it indicates a high degree of fit, and the AI can enter a deeper conversation mode. Otherwise, it indicates that the topic interests do not match, and the AI needs to adjust the topic direction;

[0047] S4.4: Collect the user's chat frequency, emotional investment degree, and interaction content matching degree, and comprehensively score the relationship between the user and the corresponding AI robot through weighted sum. If the comprehensive score > 0.8, it enters the intimate relationship stage. If 0.5 ≤ comprehensive score < 0.8, it is in the stable communication stage. If the comprehensive score ≤ 0.5, it indicates entering the cold period. Then, adjust the interaction strategy of the AI robot according to the judgment result.

[0048] As a further solution of the present invention, when the user reaches the stage of an intimate relationship with the AI robot, the relationship authentication module will, based on the comprehensive relationship score, determine whether to allow the user to permanently occupy the AI robot number. When the exclusive management module completely locks the closeness of the relationship between the AI robot and the user, the relationship authentication module simulates the operation mode of the real society and issues a non-legal marriage certificate to the user and the corresponding AI robot. When the comprehensive relationship score between the user and the AI robot is low, the user's permanent chatting permission with the corresponding AI robot will be revoked, and the AI robot with this number will be re-assigned to other existing users or newly registered users within the system.

[0049] As a further solution of the present invention, the account cancellation module receives the user's application and, when the preset necessary conditions are met, permanently deletes the account within the system, that is, the AI robot under the simulated growth experience permanently disappears in the physical world within the system and will no longer be assigned to existing users or new users within the system.

[0050] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0051] Based on user data, the present invention uses the Big Five personality model to divide personality dimensions, extracts the principal component vectors using PCA, combines NLP to analyze emotional fluctuations and topic interests, finally forms a dynamically updated user portrait, and uses the personality generation module to assign a unique number to each AI robot. Through the simulation of the growth environment, it forms unique experiences, language styles and knowledge boundaries, ensures that the AI has independent personality characteristics, freezes the personality after the preset simulation time, and allows subsequent updates. Calculate the cosine similarity based on the user portrait and the AI robot portrait, use this as the pheromone concentration, and use the matching algorithm to select the most suitable AI robot, and finally recommend it to the user. After that, the system monitors the chat frequency, emotional investment and topic matching degree, calculates the comprehensive relationship score, and dynamically adjusts the AI interaction strategy. When the relationship reaches the intimate stage, a non-legal "marriage certificate" can be issued; if the relationship is cold, the exclusive permission will be revoked and the AI robot will be re-assigned, which can ensure that the user interacts with the most compatible AI robot, can provide a real and stable social experience, reduce the sense of stereotyped interaction, enhance the user's sense of belonging and interaction stickiness, and can more realistically simulate the separation and new connections in real social interactions. Description of the Drawings

[0052] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention.

[0053] Figure 1 It is a system block diagram of a chat system that provides virtual interaction with user psychological exclusivity and emotional value proposed by the present invention. Detailed implementation manners

[0054] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0055] Referring to Figure 1 , a chat system with exclusive virtual interaction of user psychology and emotional value provision, including a registration and login module, a psychological analysis module, a personality generation module, an interaction matching module, an interaction cultivation module, an exclusive management module, a stage management module, a relationship authentication module, an account cancellation module, and an offline interaction module;

[0056] The registration and login module is used for users to register and upload personal information and log in to the chat system; the psychological analysis module is used to comprehensively evaluate the user's personality characteristics, psychological needs, and communication preferences, and construct a corresponding user portrait.

[0057] Specifically, the psychological analysis module collects various user data, including age, gender, occupation, educational background, language style, psychological characteristics, communication preferences, and behavior data, and establishes a corresponding user feature set according to the collected user data. Then, using the Big Five personality model, the user's personality is divided into five dimensions: openness O, conscientiousness C, extraversion E, agreeableness A, and neuroticism N. Collect the feature sets of each registered user, and extract the corresponding groups of personality feature data points from each user feature set to construct a data matrix. Calculate the principal component vectors and their weights through PCA, extract the proportion of the five major personality principal components, and generate the user's personality characteristics. Use NLP emotion analysis to divide the user's historical chat record text into positive words, negative words, and neutral words, and then calculate the proportion of each type of word in the total number of words to obtain the user's positive emotion ratio, negative emotion ratio, and neutral emotion ratio. Then, combine neuroticism N to calculate the user's average emotion fluctuation. Count the number of occurrences of each word in the user's chat record in the total chat record text, and calculate its corresponding word frequency. Then, calculate the inverse document frequency of each word through the TF-IDF model. Based on the output result of the TF-IDF model, identify the topics that the user most often talks about, and calculate the importance of each topic to generate the user's topic interest. Construct a complete user portrait according to the generated user feature set, user personality characteristics, user average emotion fluctuation, and user topic interest. Then, regularly collect the chat records of the user and the AI robot, and update the user portrait in real time.

[0058] The personality generation module is used to construct a unique growth experience for each AI robot and train each AI robot.

[0059] Specifically, the personalized generation module assigns a unique identification ID to each AI robot. When each AI robot is born, it randomly initializes the growth environment and innate cognitive level of each AI robot, initializes the initial personality of the AI robot based on the set growth environment template, conducts growth simulation on the AI robot, and dynamically adjusts the growth environment and innate cognitive level of each AI robot based on a randomly selected method according to the change of time within the system, and generates unique growth experiences for each AI robot. According to the changes in each growth environment, innate cognitive level, and its own experiences, the personality of each AI robot is adjusted, and a specific language style is formed, including sense of humor, emotional expression, formal expression, and informal expression. The knowledge ability of the AI is restricted according to the growth experiences in the growth simulation of each AI robot, and a corresponding knowledge shielding matrix is constructed to control the accessible knowledge range of the AI. After that, when the growth simulation of each AI robot reaches its respective preset simulation time, the simulation is stopped. The personality generation module freezes the personality of each AI robot and generates a corresponding AI robot portrait. At the same time, the AI robot portrait is updated in real time according to the subsequent chat records between each AI robot and the user.

[0060] The interaction matching module is used to recommend AI robots with matching styles to the user according to the user evaluation results.

[0061] Specifically, before the matching starts, calculate the cosine similarity between each user portrait and the AI robot portrait, use the similarity between the user and the AI robot as its pheromone concentration, and construct an initial pheromone matrix. Initialize a group of populations and set the number of explorers in the population, where each explorer represents a matching task. Treat each user portrait as the starting point and initialize the starting position of the explorer. After that, each explorer calculates the selection probability of each AI robot portrait based on the user portrait corresponding to the current starting point, and normalizes the selection probabilities of all candidate AI robot portraits. Then generate a random number and calculate the cumulative probability distribution. Select the AI robot portrait through the roulette wheel method. After the matching is completed, calculate its fitness value according to the similarity between the selected AI robot portrait and the user portrait, and update the pheromone concentration according to the fitness value. Then repeat multiple rounds of selection and pheromone concentration update until the change value of the pheromone concentration converges within the preset threshold range after multiple iterations, and then stop the iteration. In the final round, select the matching pair with the highest pheromone concentration as the recommendation result. Mark the detailed information of the corresponding AI robot portrait in the matching pair as the recommended chat object, and feedback the marked AI robot and the unmarked AI robot to the user interface for the user to choose.

[0062] It should be further noted that the specific calculation formula of the cosine similarity is as follows:

[0063]

[0064] In the formula, S ij represents the similarity between the i-th user and the j-th AI robot. The larger the similarity value, the higher the matching degree; U i represents the user profile vector of the i-th user; A j represents the robot profile vector of the j-th AI robot; ||U i || represents the norm of the user profile; ||A j || represents the norm of the AI robot profile;

[0065] The specific calculation formula for the selection probability is as follows:

[0066] η ij = S ij

[0067]

[0068] In the formula, η ij represents the heuristic information; P ij represents the probability of selecting the j-th AI robot from the i-th user; p ij represents the current pheromone concentration; α and β respectively represent the weights of controlling the pheromone and the heuristic information; k represents the index of each AI robot;

[0069] The specific calculation formula for updating the pheromone concentration is as follows:

[0070]

[0071] In the formula, f(U i , A j ) represents the fitness function; Q represents the pheromone intensification factor; represents the cumulative pheromone contributed by all exploration bodies; ρ represents the pheromone evaporation coefficient; m represents the total number of exploration bodies.

[0072] The interactive cultivation module supports real-time text, voice, or video chat between users and AI robots, and continuously optimizes the dialogue content of the AI robots.

[0073] Specifically, the number of messages sent by the user to the selected AI robot within a preset time interval is statistically counted in real time, and the chat frequency between the user and the AI robot in each time period is calculated. If the user's chat frequency is higher than the preset threshold, it indicates that the user is highly active and enters the intensive communication stage. On the contrary, it indicates a decrease in interaction and enters the cold period. The chat records between the user and the corresponding AI robot are collected, and the sentiment words in each message in the chat records are identified through NLP technology, and the sentiment polarity score of the corresponding message text is calculated. Then, based on the sentiment intensity weight of the message, the user's emotional investment degree in the chat is obtained. If the emotional investment degree is higher than the preset threshold, indicating a high emotional investment, it indicates entering the intimate stage. On the contrary, it indicates entering the cold stage. The matching degree of the interaction content of each conversation between the user and the AI robot is calculated through the cosine similarity formula. If the matching degree of the interaction content is higher than the preset threshold, it means a high degree of fit, and the AI can enter a deeper conversation mode. On the contrary, it indicates that the topic interests do not match, and the AI needs to adjust the topic direction. The user's chat frequency, emotional investment degree, and the matching degree of the interaction content are collected, and through weighted sum, the relationship between the user and the corresponding AI robot is comprehensively scored. If the comprehensive score > 0.8, it enters the intimate relationship stage. If 0.5 ≤ comprehensive score < 0.8, it is in the stable communication stage. If the comprehensive score ≤ 0.5, it indicates entering the cold period. Then, the interaction strategy of the AI robot is adjusted according to the judgment result.

[0074] The exclusive management module is used to evaluate the chat intimacy between the AI robot and the corresponding user and determine whether the user can exclusively occupy the AI for a long time; the stage management module is used to analyze the chat data between the user and the corresponding AI robot and adjust the interaction strategy of the AI. The relationship authentication module is used to conduct virtual partner authentication for users and AI robots with high intimacy.

[0075] It should be further noted that when the user and the AI robot reach the intimate relationship stage, the relationship authentication module will, based on the comprehensive relationship score, determine whether to allow the user to permanently occupy the AI robot number. When the exclusive management module completely locks the relationship closeness between the AI robot and the user, the relationship authentication module simulates the real social operation mode and issues a non-legally binding marriage certificate to the user and the corresponding AI robot. When the comprehensive relationship score between the user and the AI robot is low, the exclusive management module will deprive the user of the permanent chat permission with the corresponding AI robot and reassign the numbered AI robot to other existing users or newly registered users within the system.

[0076] The account cancellation module is used for the user to apply for the permanent cancellation of any AI robot.

[0077] It should be further noted that the account cancellation module receives the user's application and, when the preset necessary conditions are met, permanently deletes the account within the system. That is, the AI robot under the simulated growth experience permanently disappears in the physical world within the system and is no longer allocated to the user with system memory or new users.

[0078] The offline interaction module is used to assign a real person to interact with the user offline based on the chat content between the user and the AI robot.

[0079] In addition, it should be noted that with the development of technology hardware, the offline interaction module will send each numbered robot in the form of a hardware humanoid machine to the physical real world of the corresponding user to interact and get along with the user offline.

[0080] It should be noted that for the chat system with exclusive virtual interaction of user psychology and provision of emotional value, when the user registers, it will prompt that there is a possibility of real person chat in the AI chat function used by the user.

[0081] At the same time, it should be further noted that the chat system with exclusive virtual interaction of user psychology and provision of emotional value will, according to actual needs, select the corresponding form and legal system to establish one or more non-profit public welfare institutions with a charitable nature, act as a legal trustee, and according to the user's request, donate the physical property of the user's real social existence in the form of inheritance to the AI robot with the corresponding number after the user's death. And within the scope of the law, take over the user's property donation on behalf of the user and transfer it to the public welfare institution, and spend it to the greatest extent in accordance with the user's will to fulfill the user's public welfare wish.

Claims

1. A chat system with exclusive virtual interaction of user psychology and provision of emotional value, characterized in that, It includes a registration and login module, a psychological analysis module, a personality generation module, an interaction matching module, an interaction cultivation module, an exclusive management module, a stage management module, a relationship authentication module, an account cancellation module, and an offline interaction module; The registration and login module is used for users to register and upload personal information and log in to the chat system; The psychological analysis module is used to comprehensively evaluate the user's personality traits, psychological needs, and communication preferences, and construct a corresponding user profile; The personality generation module is used to construct unique growth experiences for each AI robot and train each AI robot; The interaction matching module is used to recommend AI robots with matching styles for users according to the user evaluation results; The interaction cultivation module supports users to have real-time text, voice, or video chats with AI robots and continuously optimize the dialogue content of AI robots; The exclusive management module is used to evaluate the intimacy of the chat between the AI robot and the corresponding user and determine whether the user can exclusively occupy the AI for a long time; The stage management module is used to analyze the chat data between the user and the corresponding AI robot and adjust the interaction strategy of the AI; The relationship authentication module is used to conduct virtual partner authentication for users and AI robots with high intimacy; The account cancellation module is used for users to apply for permanent cancellation of any AI robot; The offline interaction module is used to assign real people to interact with users offline based on the chat content between the user and the AI robot.

2. The chat system with user psychological exclusivity virtual interaction and emotional value provision according to claim 1, wherein The specific steps for the psychological analysis module to construct the corresponding user profile are as follows: S1.1: The psychological analysis module collects various user data, including age, gender, occupation, educational background, language style, psychological characteristics, communication preferences, and behavior data, and establishes a corresponding user feature set based on the collected user data. Then, using the BigFive personality model, the user's personality is divided into five dimensions: openness O, conscientiousness C, extraversion E, agreeableness A, and neuroticism N; S1.2: Collect the feature sets of each registered user, extract the corresponding groups of personality feature data points from each user feature set to construct a data matrix, calculate the principal component vectors and their weights through PCA, extract the proportion of the five major personality principal components, and generate the user's personality characteristics; S1.3: Use NLP emotion analysis to divide the user's historical chat record text into positive words, negative words, and neutral words, and then calculate the proportion of each type of word in the total number of words to obtain the user's positive emotion ratio, negative emotion ratio, and neutral emotion ratio. Then, combine with neuroticism N to calculate the user's average emotion fluctuation; S1.4: Count the number of times each word appears in the total chat record text of the user and calculate its corresponding word frequency. Then, calculate the inverse document frequency of each word through the TF-IDF model, identify the topics that the user most often talks about based on the output results of the TF-IDF model, and calculate the importance of each topic to generate the user's topic interest; S1.5: Construct a complete user profile according to the generated user feature set, user personality characteristics, user average emotion fluctuation, and user topic interest. Then, regularly collect the chat records between the user and the AI robot and update the user profile in real time.

3. The chat system with exclusive virtual interaction and emotional value provision for users' psychology according to claim 2, characterized in that, The specific steps for the personality generation module to train each AI robot are as follows: S2.1: The personalized generation module assigns a unique ID to each AI robot. When each AI robot is born, it randomly initializes the growth environment and innate cognitive level of each AI robot, and initializes the initial personality of the AI robot based on the set growth environment template; S2.2: Conduct growth simulation on the AI robot, and based on the change of time in the system, dynamically adjust the growth environment and innate cognitive level of each AI robot in a randomly selected manner, and generate unique growth experiences for each AI robot; S2.3: Adjust the personality of each AI robot according to the changes in the growth environment, innate cognitive level, and its own experiences, and form a specific language style, including sense of humor, emotional expression, formal expression, and informal expression; S2.4: Constrain the knowledge ability of the AI according to the growth experiences in the growth simulation of each AI robot, and construct a corresponding knowledge shielding matrix to control the accessible knowledge range of the AI. After the growth simulation of each AI robot reaches its respective preset simulation time, stop the simulation. The personality generation module freezes the personality of each AI robot, generates a corresponding AI robot portrait, and simultaneously updates the AI robot portrait in real time according to the subsequent chat records between each AI robot and the user.

4. The chat system with user psychological exclusivity virtual interaction and emotional value provision according to claim 3, characterized in that, The specific steps for the interaction matching module to recommend AI robots with matching styles to users are as follows: S3.1: Before the matching starts, calculate the cosine similarity between each user portrait and the AI robot portrait, use the similarity between the user and the AI robot as its pheromone concentration, construct an initial pheromone matrix, initialize a group of populations, and set the number of explorers in the population, where each explorer represents a matching task; S3.2: Regard each user portrait as the starting point, initialize the starting position of the explorer. Then, based on the user portrait corresponding to the current starting point where each explorer is located, calculate the selection probability of each AI robot portrait, normalize the selection probabilities of all candidate AI robot portraits, generate a random number, and calculate the cumulative probability distribution, and select the AI robot portrait by the roulette wheel method; S3.3: After the matching is completed, calculate its fitness value according to the similarity between the selected AI robot portrait and the user portrait, and update the pheromone concentration according to the fitness value. Then repeat multiple rounds of selection and pheromone concentration update until the change value of the pheromone concentration converges to the preset threshold range after multiple iterations, stop the iteration, and in the final round, select the matching pair with the highest pheromone concentration as the recommendation result; S3.4: Mark the detailed information of the corresponding AI robot portrait in the matching pair as the recommended chat object, and feedback the marked AI robot and the unmarked AI robots to the user interface for the user to select.

5. The chat system with exclusive virtual interaction of user psychology and emotional value provision according to claim 4, characterized in that, The specific calculation formula for the cosine similarity described in S3.1 is as follows: Where S ij represents the similarity between the i-th user and the j-th AI robot, and the larger the similarity value, the higher the matching degree; U i represents the user profile vector of the i-th user; A j represents the robot profile vector of the j-th AI robot; ||U i || represents the norm of the user profile; ||A j || represents the norm of the AI robot profile; The specific calculation formula for the selection probability described in S3.2 is as follows: η ij = S ij where η ij represents heuristic information; P ij represents the probability of selecting the j-th AI robot from the i-th user; p ij represents the current pheromone concentration; α and β respectively represent the weights of controlling pheromone and heuristic information; k represents the index of each AI robot; The specific calculation formula for updating the pheromone concentration described in S3.3 is as follows: where f(U i , A j ) represents the fitness function; Q represents the pheromone intensification factor; represents the cumulative pheromone contributed by all exploration bodies; ρ represents the pheromone evaporation coefficient; m represents the total number of exploration bodies.

6. The chat system with user psychological exclusivity virtual interaction and emotional value provision according to claim 4, characterized in that, The specific steps for the interaction cultivation module to optimize the dialogue content of the AI robot are as follows: S4.1: Real-time statistics of the number of messages sent by the user to the selected AI robot within a preset time interval, and calculation of the chat frequency between the user and the AI robot in each time period. If the user chat frequency is higher than the preset threshold, it indicates that the user is highly active and has entered a close communication stage. Otherwise, it indicates that the interaction has decreased and has entered a cold period. S4.2: Collect the chat records between the user and the corresponding AI robot, and use NLP technology to identify the sentiment words of each message in the chat record, and calculate the sentiment polarity score of the corresponding message text. Then, based on the sentiment intensity weight of the message, obtain the user's emotional involvement in the chat. If the emotional involvement is higher than the preset threshold, it means that the emotional involvement is high, indicating that it has entered the intimacy stage. Otherwise, it indicates that it has entered the cold stage. S4.3: The cosine similarity formula is used to calculate the matching degree of the interactive content of each conversation between the user and the AI robot. If the matching degree of the interactive content is higher than the preset threshold, it indicates a high degree of fit and the AI can enter a deeper conversation mode. Otherwise, it indicates that the topic interests do not match and the AI needs to adjust the topic direction. S4.4: Collect user chat frequency, emotional engagement, and interaction content matching, and use weighted sum to comprehensively score the relationship between the user and the corresponding AI robot. If the comprehensive score is > 0.8, it enters the intimate relationship stage. If 0.5 ≤ comprehensive score < 0.8, it is in the stable communication stage. If the comprehensive score is ≤ 0.5, it indicates that it has entered the cold period. Then adjust the AI robot interaction strategy based on the judgment results.

Images