Virtual digital human interaction management method and system based on artificial intelligence

Abstract

The invention relates to the technical field of data exchange, in particular to a virtual digital human interaction management method and system based on artificial intelligence. Comprising the following steps: S1, establishing a portrait identity database, and independently storing a consultation record of each user according to portrait data; s2, collecting historical business consultation data of a bank, and then carrying out statistical analysis on the historical business consultation data by utilizing artificial intelligence so as to obtain the number of first consultation times of each business type; through different processing modes for new users and old users, services can be provided for the users more efficiently, input voices of the new users are analyzed in combination with historical consultation service data, service matching is performed in combination with the number of times of first consultation, accurate information of related services can be provided for the new users quickly, and user experience is improved. And the old user can carry out business matching according to the historical consultation record in the association effective time threshold value and the business type associated with the consultation sequence, so that more personalized services are provided.

Description

Technical Field

[0001] The present invention relates to the technical field of data exchange, and in particular to a method and system for managing virtual digital human interaction based on artificial intelligence. Background Art

[0002] With the rapid development of artificial intelligence technology, the field of banking business consulting is also constantly exploring innovative service methods. In banking business consulting, services are mainly provided to users through traditional manual customer service, self-service inquiry terminals, etc., to help users solve various problems.

[0003] At present, in bank branches, users can obtain business information by communicating with staff face to face or using self-service inquiry terminals. On the online customer service platform, users can communicate with customer service staff through text chat, telephone, etc. However, the service efficiency of manual customer service is low, and it is difficult to meet the simultaneous consultation needs of a large number of users, which easily leads to long waiting times for users. Secondly, traditional self-service inquiry terminals have relatively single functions and cannot provide personalized services, making it difficult to meet the diverse needs of users. These defects have a great impact on achieving the purpose of efficient and personalized banking business consultation. Therefore, a virtual digital human interaction management method and system based on artificial intelligence are proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a virtual digital human interaction management method and system based on artificial intelligence to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, one of the purposes of the present invention is to provide a virtual digital human interaction management method based on artificial intelligence, comprising the following steps:

[0006] S1. Establish a portrait identity database and store each user's consultation record separately according to the portrait data;

[0007] S2. Collect the bank's historical business consultation data, and then use artificial intelligence to perform statistical analysis on the historical business consultation data to obtain the number of first consultations for each business type, and perform consultation sequence correlation analysis between business types;

[0008] S3, collecting the portrait data of the consulting service user through the banking service consulting device, and collecting the user's input voice at the same time, and then identifying the portrait data in the database, and performing S4 and S5 according to the identification results;

[0009] S4. When the user is identified as a new user, the input voice is combined with the historical consultation business data for relevant business analysis, and then the analysis result is combined with the number of first consultations for business matching, and then the virtual digital human generates response content according to the matched business;

[0010] S5. When the recognition shows that the user is an old user, the input voice is combined with the historical consultation business data for relevant business analysis, and the associated effective time threshold is set. Then, the business types associated with the historical consultation records and consultation sequence within the associated effective time threshold and the analysis results are matched with the business. After that, the virtual digital person generates response content according to the matched business.

[0011] As a further improvement of the present technical solution, S1 establishes a portrait identity database in the banking business consulting device, obtains portrait data through the camera of the banking business consulting device, and then stores the consulting records of each user separately according to the portrait data.

[0012] As a further improvement of the technical solution, the steps of S2 are as follows:

[0013] S2.1, obtain the historical business consultation data of the bank counter, then aggregate the consultation records stored in S1 and the historical business consultation data of the bank counter, and then use artificial intelligence to perform statistical analysis to obtain the number of first consultations for each business type;

[0014] S2.2. Conduct consultation sequence correlation analysis between business types to obtain other business types with the most repeated consultations after consulting each business type.

[0015] As a further improvement of the present technical solution, the S3 sets up an identity authentication module. When a user needs to conduct business consultation through the banking business consultation device, the user needs to first conduct identity authentication through portrait collection before the business consultation can be conducted. At the same time, a virtual digital person is set up in the banking business consultation device, and the virtual digital person interacts with the user to complete the user's business consultation needs.

[0016] As a further improvement of the technical solution, the steps of S3 are as follows:

[0017] S3.1. After the user passes identity authentication, the microphone of the banking consulting device collects the user's input voice;

[0018] S3.2. The collected portrait data is combined with the portrait identity database for authentication analysis. When the portrait data matches the corresponding user in the portrait identity database, the user is determined to be an old user and the collected voice is sent to S5. Conversely, when the portrait data does not match the corresponding user in the portrait identity database, the user is determined to be a new user and the collected input voice is sent to S4.

[0019] As a further improvement of the technical solution, the step of S4 is as follows:

[0020] S4.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S4.2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value;

[0021] S4.2. The accurate value of each business is combined with the number of first consultations for each business type to perform business matching, thereby comparing the businesses with accurate values ​​in combination with the business types, obtaining the business type with the highest proportion of total accurate values, and then selecting the business with the highest accurate value in the business type to generate the response content.

[0022] As a further improvement of the technical solution, the step of S5 is as follows:

[0023] S5.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S5.2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value;

[0024] S5.2. Set an associated effective time threshold, extract the user's historical consultation records at the same time, filter the historical consultation records in combination with the effective time threshold, obtain effective consultation records within the associated effective time threshold, and then extract the business type that has been consulted most recently from the effective consultation records, and then obtain other business types with the most repeated consultations based on the most recent business type, and then select the business with the highest accuracy value based on other business types to generate response content.

[0025] As a further improvement of the present technical solution, S4 and S5 generate response content and feed it back to the banking business consulting device, and then the virtual digital person selects a presentation method according to the response content, and then provides business consulting solutions to the user according to the selected presentation method.

[0026] The second object of the present invention is to provide a virtual digital human interaction management system based on artificial intelligence, including any one of the above-mentioned virtual digital human interaction management methods based on artificial intelligence, including a database establishment unit, a portrait recognition unit and a response content generation unit;

[0027] The database establishment unit is used to establish a portrait identity database, collect the bank's historical business consultation data, and then use artificial intelligence to perform statistical analysis on the historical business consultation data, so as to obtain the number of first consultations for each business type, and at the same time perform consultation sequence correlation analysis between business types;

[0028] The portrait recognition unit is used to collect the portrait data of the consulting service user through the banking service consulting device, and collect the user's input voice at the same time, and then recognize the portrait data in the database;

[0029] The response content generation unit is used to combine the input voice with the historical consulting business data to perform relevant business analysis, and then perform business matching, after which the virtual digital person generates response content according to the matched business.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. A virtual digital human interaction management method and system based on artificial intelligence, which can provide services to users more efficiently through different processing methods for new users and old users. The input voice of new users is analyzed in combination with historical consultation business data, and business matching is performed based on the number of first consultations, which can quickly provide new users with accurate information on related businesses. For old users, business matching can be performed based on historical consultation records within the associated effective time threshold and business types associated with the consultation sequence, providing more personalized services. At the same time, through the calculation of accurate values ​​and threshold comparison, as well as the analysis of the proportion of business types and the selection of the business with the highest accurate value, high-quality response content can be generated to ensure that the answers obtained by users are accurate.

[0032] 2. A virtual digital human interaction management method and system based on artificial intelligence. By statistically analyzing the bank's historical business consultation data, the number of first consultations for each business type and the consultation sequence correlation between business types are obtained, which helps the bank understand the popularity of different businesses and the trend of user consultations, and helps the bank optimize its business layout and resource allocation to better meet user needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is the overall flow chart of the present invention;

[0034] Figure 2 A flowchart of obtaining the number of first consultations for each business type in the present invention;

[0035] Figure 3 A flowchart of collecting user input speech of the present invention;

[0036] Figure 4 A flowchart of generating response content for a service with the highest accuracy value according to the present invention;

[0037] Figure 5 A flowchart of the present invention for selecting the service with the highest accuracy value according to other service types to generate response content;

[0038] Figure 6 This is a structural principle diagram of the database establishment unit of the present invention.

[0039] The meaning of each number in the figure is:

[0040] 10. Database establishment unit; 20. Portrait recognition unit; 30. Response content generation unit. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1 - Figure 6 As shown, one of the purposes of the present invention is to provide a virtual digital human interaction management method based on artificial intelligence, comprising the following steps:

[0043] S1. Establish a portrait identity database and store each user's consultation record separately according to the portrait data;

[0044] S1 establishes a portrait identity database in the banking business consulting device, obtains portrait data through the camera of the banking business consulting device, and then stores the consultation records of each user separately according to the portrait data. The specific steps are as follows:

[0045] Establish a portrait identity database: Install a high-resolution camera on the banking consulting device to ensure that the user's portrait can be clearly captured, design a user interaction interface to guide users to collect their portraits when using the device, and store the extracted portrait feature vectors in the database, which can be associated with other user information, such as the user's name;

[0046] Acquiring portrait data: When a user approaches the banking consultation device, the camera is automatically activated to continuously capture multiple frames of portrait images to ensure that a clear and stable portrait is obtained;

[0047] The consultation record of each user is stored separately based on the portrait data: the portrait feature vector extracted in real time is compared with the feature vector in the portrait identity database to find the most similar portrait. Once a matching portrait is found, the user's consultation record is associated with the user's ID or other identification information and stored in the database.

[0048] S2. Collect the bank's historical business consultation data, and then use artificial intelligence to perform statistical analysis on the historical business consultation data to obtain the number of first consultations for each business type, and perform consultation sequence correlation analysis between business types;

[0049] The steps of S2 are as follows:

[0050] S2.1. Obtain the historical business consultation data of the bank counter, then aggregate the consultation records stored in S1 and the historical business consultation data of the bank counter, and then use artificial intelligence to perform statistical analysis to obtain the number of first consultations for each business type. The specific steps are as follows:

[0051] Determine the data source: connect with the bank's database management system, determine the database table that stores historical business consulting data, and use database query languages ​​such as SQL or data extraction tools to extract historical business consulting data from the data source;

[0052] Data merging: Merge the historical business consultation data obtained from the bank counter with the consultation records previously stored in S1, and store the merged data in a new database table for subsequent analysis;

[0053] Use artificial intelligence for statistical analysis: pre-process the merged consultation data, including encoding business types, extracting time information, etc. For each business type, count the number of first consultations.

[0054] S2.2. Conduct consultation sequence correlation analysis between business types to obtain the other business types with the most repeated consultations after consulting each business type. The formula is as follows:

[0055]

[0056] Among them, A and B are different business types, n(A) is the number of consultation records of business type A, n(A∪B) is the number of consultation records of business types A and B appearing at the same time, that is, the situation where A is consulted first and then B, and Confidence(A→B) is the probability of consulting business type B after consulting business type A.

[0057] S3, collecting the portrait data of the consulting service user through the banking service consulting device, and collecting the user's input voice at the same time, and then identifying the portrait data in the database, and performing S4 and S5 according to the identification results;

[0058] S3 sets up an identity verification module. When a user needs to conduct business consultation through the banking business consultation device, he / she needs to first conduct identity verification through face image collection before he / she can conduct business consultation. At the same time, a virtual digital person is set up in the banking business consultation device to interact with the user and complete the user's business consultation needs.

[0059] When the user needs to make a business consultation, the identity authentication module is started, the camera collects the user's portrait, and performs real-time feature extraction. The real-time extracted portrait feature vector is compared with the stored portrait feature, and the similarity is calculated using the distance measurement formula. If the similarity exceeds a certain threshold, the identity authentication is considered successful and the user is allowed to make a business consultation. Otherwise, the user is prompted to re-collect the portrait.

[0060] Choose a suitable virtual digital human model, which can be a pre-trained general model or customized training based on the bank's business needs. Use a large amount of text data and voice data to train the virtual digital human to improve its language understanding and generation capabilities, as well as speech synthesis quality.

[0061] The steps for S3 are as follows:

[0062] S3.1. After the user passes identity authentication, the microphone of the banking consulting device collects the user's input voice;

[0063] S3.2, the collected portrait data is combined with the portrait identity database for authentication analysis. When the portrait data matches the corresponding user in the portrait identity database, the user is determined to be an old user, and the collected voice is sent to S5. On the contrary, when the portrait data does not match the corresponding user in the portrait identity database, the user is determined to be a new user, and the collected input voice is sent to S4. The specific steps are as follows:

[0064] Voice collection: Ensure that the microphone of the banking consulting device is turned on normally and initialized, so that it can collect the user's input voice at any time. When the user passes the identity authentication, the system automatically activates the microphone to start collecting the user's input voice information;

[0065] Database comparison: compare and analyze the acquired portrait data with the established portrait identity database;

[0066] When it is determined that the user is an old user, that is, the portrait data matches the corresponding user in the portrait identity database, the collected user input voice data is sent to S5 according to the established communication protocol and data transmission method;

[0067] When it is determined that the user is a new user, that is, the portrait data does not match the corresponding user in the portrait identity database, the collected input voice is also sent to S4 according to the prescribed transmission method.

[0068] S4. When the user is identified as a new user, the input voice is combined with the historical consultation business data for relevant business analysis, and then the analysis result is combined with the number of first consultations for business matching, and then the virtual digital human generates response content according to the matched business;

[0069] The steps of S4 are as follows:

[0070] S4.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S4.2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value. The specific steps are as follows;

[0071] Assume that there are n businesses in the historical consulting business data, and the feature set extracted from the i-th business is F i ={f i1 , f i2 , …, f im}, m is the number of features extracted for each service, and the input speech is converted into text, and then feature extraction is performed. Let the extracted feature set be F input ={f input1 , f input2 , …, f inputk}, k is the number of features extracted from the input speech text, and then the formula for calculating the exact value between each business and the input speech is as follows:

[0072]

[0073] Among them, A i is the exact value of the i-th service and the input voice, f input1 refers to the lth feature extracted from the text converted from the input speech, f ij is the j-th feature of the i-th business from the historical consulting business data.

[0074] S4.2. The accurate value of each business is combined with the number of first consultations for each business type to perform business matching, thereby comparing the businesses with accurate values ​​in combination with the business types, obtaining the business type with the highest proportion of total accurate values, and then selecting the business with the highest accurate value in the business type to generate the response content.

[0075] S5. When the recognition shows that the user is an old user, the input voice is combined with the historical consultation business data for relevant business analysis, and the associated effective time threshold is set. Then, the business types associated with the historical consultation records and consultation sequence within the associated effective time threshold and the analysis results are matched with the business. After that, the virtual digital person generates response content according to the matched business.

[0076] The steps of S5 are as follows;

[0077] S5.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S5.2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value;

[0078] Same as the method in S4.1 above

[0079] S5.2. Set the association validity time threshold, extract the user's historical consultation records at the same time, filter the historical consultation records in combination with the validity time threshold, obtain the valid consultation records within the association validity time threshold, and then extract the business type with the most recent consultation from the valid consultation records, and then obtain other business types with the most re-consultations based on the most recent consultation business type, and then select the business with the highest accuracy value based on other business types to generate response content. The specific steps are as follows:

[0080] Set association validity time threshold: Determine a reasonable association validity time threshold, such as setting it to the past month;

[0081] Filter valid consultation records: traverse the extracted historical consultation records, and for each record, calculate the time difference between the current time and the consultation time of the record, and then compare the time difference with the valid time threshold. If the time difference is within the threshold range, mark the record as a valid consultation record and retain it;

[0082] Extract the business type of the most recent consultation: sort the filtered valid consultation records by consultation time, ensure that the latest consultation record is in the front, then take the first record and extract the business type therein, which is the business type of the most recent consultation;

[0083] Obtain other business types with the most re-consultations: Count the number of times each business type appears after the business type with the most recent consultation in the valid consultation records, then compare the number of times different business types appear, and determine the business type with the most appearances, which is the other business type with the most re-consultations;

[0084] Select the business with the highest accuracy to generate the response content: among the business types with the most repeated consultations, find the accurate value of each business, compare the accurate values ​​of these businesses, and find the business with the highest accuracy to generate the response content, so that the virtual digital human can provide answers to the user's business consultations.

[0085] S4 and S5 generate response content and feed it back to the banking business consultation device. Then the virtual digital human selects a presentation method based on the response content and then provides business consultation solutions to the user based on the selected presentation method. The specific steps are as follows:

[0086] Feedback response content: Generate response content according to the previous S4.2 and S5.2, and send the response content to the banking business consulting device;

[0087] The virtual digital human selects a presentation method: a presentation method is selected based on the characteristics of the response content, user preferences, and device capabilities;

[0088] Once the presentation method is determined, the virtual digital human uses this method to present the response content to the user;

[0089] If it is a voice presentation, use speech synthesis technology to convert the response content into speech and play it back;

[0090] If it is a text presentation, display the response content on the device's display screen;

[0091] If it is an image presentation, generate the corresponding image and display it to the user.

[0092] The second object of the present invention is to provide a virtual digital human interaction management system based on artificial intelligence, including any one of the above-mentioned virtual digital human interaction management methods based on artificial intelligence, including a database establishment unit 10, a portrait recognition unit 20 and a response content generation unit 30;

[0093] The database establishment unit 10 is used to establish a portrait identity database, collect the historical business consultation data of the bank, and then use artificial intelligence to perform statistical analysis on the historical business consultation data, so as to obtain the number of first consultations for each business type, and at the same time perform consultation sequence correlation analysis between business types;

[0094] The portrait recognition unit 20 is used to collect the portrait data of the consulting service user through the banking service consulting device, and collect the user's input voice at the same time, and then recognize the portrait data in the database;

[0095] The response content generation unit 30 is used to combine the input voice with the historical consulting business data to perform relevant business analysis, and then perform business matching, after which the virtual digital human generates response content based on the matched business.

[0096] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A virtual digital human interaction management method based on artificial intelligence, characterized in that: The steps include: S1. Establish a portrait identity database and store each user's consultation record separately according to the portrait data; S2. Collect the bank's historical business consultation data, and then use artificial intelligence to perform statistical analysis on the historical business consultation data to obtain the number of first consultations for each business type, and perform consultation sequence correlation analysis between business types; S3, collecting the portrait data of the consulting service user through the banking service consulting device, and collecting the user's input voice at the same time, and then identifying the portrait data in the database, and performing S4 and S5 according to the identification results; S4. When the user is identified as a new user, the input voice is combined with the historical consultation business data for relevant business analysis, and then the analysis result is combined with the number of first consultations for business matching, and then the virtual digital human generates response content according to the matched business; S5. When the recognition shows that the user is an old user, the input voice is combined with the historical consultation business data for relevant business analysis, and the associated effective time threshold is set. Then, the business types associated with the historical consultation records and consultation sequence within the associated effective time threshold and the analysis results are matched with the business. After that, the virtual digital person generates response content according to the matched business.

2. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The S1 establishes a portrait identity database in the banking business consulting device, obtains portrait data through the camera of the banking business consulting device, and then stores the consulting records of each user separately according to the portrait data.

3. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The steps of S2 are as follows: S2.1, obtain the historical business consultation data of the bank counter, then aggregate the consultation records stored in S1 and the historical business consultation data of the bank counter, and then use artificial intelligence to perform statistical analysis to obtain the number of first consultations for each business type; S2.

2. Conduct consultation sequence correlation analysis between business types to obtain other business types with the most repeated consultations after consulting each business type.

4. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The S3 sets up an identity authentication module. When a user needs to conduct business consultation through the bank business consultation device, the user needs to first conduct identity authentication through portrait collection before the business consultation can be conducted. At the same time, a virtual digital person is set up in the bank business consultation device, and the virtual digital person interacts with the user to complete the user's business consultation needs.

5. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The steps of S3 are as follows: S3.

1. After the user passes identity authentication, the microphone of the banking consulting device collects the user's input voice; S3.

2. The collected portrait data is combined with the portrait identity database for authentication analysis. When the portrait data matches the corresponding user in the portrait identity database, the user is determined to be an old user and the collected voice is sent to S5. Conversely, when the portrait data does not match the corresponding user in the portrait identity database, the user is determined to be a new user and the collected input voice is sent to S4.

6. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The steps of S4 are as follows: S4.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S4.

2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value; S4.

2. The accurate value of each business is combined with the number of first consultations for each business type to perform business matching, thereby comparing the businesses with accurate values ​​in combination with the business types, obtaining the business type with the highest proportion of total accurate values, and then selecting the business with the highest accurate value in the business type to generate the response content.

7. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: The steps of S5 are as follows: S5.1, receiving the input voice sent by S3.2, and then combining the input voice with the historical consulting business data to perform relevant business analysis, obtain the accurate value of each business and the input voice, and set the accurate value threshold at the same time, and then compare the accurate value of each business with the accurate value threshold. When the accurate value of no business exceeds the accurate value threshold, continue to S5.

2. When the accurate value of a business exceeds the accurate value threshold, generate response content with the business with the highest accurate value; S5.

2. Set an associated effective time threshold, extract the user's historical consultation records at the same time, filter the historical consultation records in combination with the effective time threshold, obtain effective consultation records within the associated effective time threshold, and then extract the business type that has been consulted most recently from the effective consultation records, and then obtain other business types with the most repeated consultations based on the most recent business type, and then select the business with the highest accuracy value based on other business types to generate response content.

8. The method for managing virtual digital human interaction based on artificial intelligence according to claim 1, characterized in that: S4 and S5 generate response content and feed it back to the banking business consultation device, and then the virtual digital person selects a presentation method according to the response content, and then provides business consultation solutions to the user according to the selected presentation method.

9. A method for implementing an artificial intelligence-based virtual digital human interaction management system, comprising an artificial intelligence-based virtual digital human interaction management method according to any one of claims 1 to 8, characterized in that: It comprises a database establishment unit (10), a human image recognition unit (20) and a response content generation unit (30); The database establishment unit (10) is used to establish a portrait identity database, collect historical business consultation data of the bank, and then use artificial intelligence to perform statistical analysis on the historical business consultation data, so as to obtain the number of first consultations for each business type, and at the same time perform consultation sequence correlation analysis between business types; The portrait recognition unit (20) is used to collect the portrait data of the consulting service user through the banking service consulting device, and collect the user's input voice at the same time, and then recognize the portrait data in the database; The response content generating unit (30) is used to combine the input voice with the historical consulting business data to perform relevant business analysis, and then perform business matching, after which the virtual digital person generates response content according to the matched business.

Images