Intelligent conversation expert system

Abstract

The invention provides an intelligent conversation expert system, and the system comprises a pre-preparation module which carries out the mining and analysis of the related data of a talked person in a database through a multi-mode large model, predicts the psychological state and behavior tendency of the talked person, and generates a conversation strategy and a preparation suggestion at a pre-conversation preparation stage; the conversation-in-event module is used for collecting image data and voice data of a talked person in real time in the conversation process, analyzing the speech, expression and action of the talked person through a multi-modal large model, predicting the psychological state and behavior tendency of the talked person and generating real-time emotion feedback information and conversation guide suggestions; and the afterwards summarizing module is used for designing the conversation content analysis module by utilizing a multi-mode large model after the conversation is finished, summarizing the conversation effect and optimizing the conversation strategy in combination with the conversation type, and continuously improving the intelligent level of the system through machine learning. The method mainly solves the problem that a conversation person in a special management place lacks data support, real-time feedback and strategy optimization in the conversation process.

Description

Technical Field

[0001] The present invention mainly relates to technical fields such as artificial intelligence, big data analysis, natural language processing and conversation management, and in particular to an intelligent conversation expert system. Background Art

[0002] The conversation between the interviewer and the person being interviewed is an important part of the management of special management places. It helps to understand the ideological trends, psychological conditions and needs of the person being interviewed, and at the same time educate and influence the person being interviewed.

[0003] In the traditional management system of special management places, the interviewer mainly relies on personal experience and intuition to prepare for the interview and formulate strategies, lacking scientific data support and real-time feedback. At present, the management of special management places is gradually developing in the direction of intelligence and informationization, but the existing intelligent systems are mostly focused on basic functions such as monitoring and alarm, and lack support for advanced functions such as interaction between the interviewer and the interviewee, psychological analysis, and strategy optimization, and cannot achieve in-depth behavior analysis of the interviewee, psychological state prediction, and real-time optimization of the interview strategy. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent conversation expert system, which mainly solves the problem that talkers in special management places lack data support, real-time feedback and strategy optimization during the conversation process.

[0005] The present invention provides an intelligent conversation expert system, comprising:

[0006] The pre-conversation preparation module uses a multimodal large model to mine and analyze the interviewee's background information and historical behavior data in the database, predict the interviewee's psychological state and behavioral tendencies, and generate conversation strategies and preparation suggestions based on the conversation type.

[0007] The in-conversation module collects the image and voice data of the person being interviewed in real time during the conversation, analyzes the person's speech, expression and actions through a multimodal large model, predicts the person's psychological state and behavioral tendencies, and generates real-time emotional feedback information and conversation guidance suggestions based on the conversation type;

[0008] The post-conversation summary module uses the learning and optimization capabilities of the multimodal large model to design a conversation content analysis module after the conversation. It summarizes the conversation effect based on the conversation type, optimizes the conversation strategy, and continuously improves the intelligence level of the system through machine learning.

[0009] One or more technical solutions provided by the present invention have at least the following technical effects or advantages: in the pre-conversation preparation stage, the in-conversation stage, and the post-conversation summary stage, the multimodal large model technology is used to conduct in-depth multi-dimensional analysis and prediction of the interviewee, and provide feedback and suggestions to the interviewer; improve the pertinence and effectiveness of the interviewer's conversation, reduce the resistance of the interviewee, and improve the management efficiency and safety of special management places. Through data-driven strategy optimization, the interviewer's conversation skills and the overall management level of special management places are continuously improved.

[0010] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0012] Figure 1 Schematic diagram of the system architecture in an embodiment of the present invention. DETAILED DESCRIPTION

[0013] The embodiment of the present application provides an intelligent conversation expert system, which mainly solves the problem that conversation persons in special management places lack data support, real-time feedback and strategy optimization during the conversation process.

[0014] The technical solution in the embodiment of the present application has the following general idea: in the pre-conversation preparation stage, the in-conversation stage, and the post-conversation summary stage, the multimodal large model technology is used to conduct in-depth multi-dimensional analysis and prediction of the interviewee, and provide feedback and suggestions to the interviewer; improve the pertinence and effectiveness of the interviewer's conversation, reduce the resistance of the interviewee, and improve the management efficiency and safety of special management places. Through data-driven strategy optimization, the interviewer's conversation skills and the overall management level of special management places are continuously improved.

[0015] A multimodal big model refers to a big model that can process multiple different types of data, such as text, images, audio and other multimodal data. This type of model combines the capabilities of NLP (language big model) and CV (visual big model) to achieve a comprehensive understanding and analysis of multimodal information, so that it can more comprehensively understand and process complex data. According to different application fields, it can be mainly divided into three levels: L0, L1, and L2:

[0016] The L0 level is a general-purpose large model, which refers to a large model that can be used in multiple fields and tasks. They use large computing power, massive open data, and deep learning algorithms with huge parameters to train on large-scale unlabeled data to find features and discover patterns, thereby forming a powerful generalization ability that can "draw inferences from one example to another". It can complete multi-scenario tasks without fine-tuning or with a small amount of fine-tuning, which is equivalent to AI completing "general education".

[0017] The L1 level, or industry-specific large models, refers to large models that target specific industries or fields. They are usually pre-trained or fine-tuned using industry-related data to improve performance and accuracy in that field, which is equivalent to AI becoming an "industry expert."

[0018] The L2 level, or vertical large models, refers to large models that are targeted at specific tasks or scenarios. They are usually pre-trained or fine-tuned using task-related data to improve performance and results on the task.

[0019] The intelligent conversation expert system of the present invention belongs to a vertical large model, which is a large model for a specific task or scenario of a conversation between a talker and a talkee. It can be obtained by pre-training or fine-tuning an existing general large model, and the pre-training or fine-tuning includes:

[0020] 1. Data collection and annotation:

[0021] Speech and emotion data of the interviewee: Collect a large amount of conversation data of the interviewee and annotate their emotions, such as happiness, anger, frustration, etc.

[0022] Conversation content classification: mark the topic and intention of the conversation, such as asking for help, regret, complaint, plan, etc.

[0023] 2. Sentiment Analysis:

[0024] Fine-tuning the emotion recognition model: Use the labeled emotion data to fine-tune the model so that it can accurately identify the emotional changes of the person being interviewed;

[0025] Multimodal sentiment analysis: Combines multiple data sources such as speech, text, and facial expressions to improve the accuracy of emotion recognition.

[0026] 3. Language comprehension and generation:

[0027] Language model fine-tuning: For special scenarios of conversations in special management places, the language model is fine-tuned so that it can understand and generate conversations related to this scenario;

[0028] Content analysis: Conduct in-depth analysis of the interviewee’s conversation content to identify key issues and potential risks.

[0029] After the general large model is pre-trained or fine-tuned through the above process, the intelligent conversation expert system of the present invention is obtained, which can be applied to conversation scenarios between the talker and the person being talked to in special management places, including prior preparation, in-conversation assistance and post-conversation summary optimization.

[0030] like Figure 1 As shown, this embodiment provides an intelligent conversation expert system, including:

[0031] The pre-conversation preparation module uses a multimodal large model to mine and analyze the interviewee's background information and historical behavior data in the database, predict the interviewee's psychological state and behavioral tendencies, and generate conversation strategies and preparation suggestions based on the conversation type.

[0032] The in-conversation module collects the image and voice data of the person being interviewed in real time during the conversation, analyzes the person's speech, expression and actions through a multimodal large model, predicts the person's psychological state and behavioral tendencies, and generates real-time emotional feedback information and conversation guidance suggestions based on the conversation type;

[0033] The post-conversation summary module uses the learning and optimization capabilities of the multimodal large model to design a conversation content analysis module after the conversation. It summarizes the conversation effect based on the conversation type, optimizes the conversation strategy, and continuously improves the intelligence level of the system through machine learning.

[0034] Furthermore, this embodiment may also include the following functional modules to provide necessary assistance to inexperienced talkers:

[0035] The knowledge base query and recommendation module is used to provide query and recommendation of professional knowledge such as laws and regulations, psychological knowledge, and conversation skills during the preparation stage before or during the conversation.

[0036] The case library retrieval module is used to retrieve relevant successful cases from the case library in real time during the preparation stage before the conversation or during the conversation, based on the conversation topic or the characteristics of the person being interviewed, to provide reference for the interviewer.

[0037] The recording and analysis module is used to automatically record the conversation process, perform content analysis, and generate conversation reports and improvement suggestions.

[0038] Furthermore, this embodiment may also include the following functional modules to facilitate the understanding and use of the speaker and improve the accuracy and effectiveness of the system:

[0039] The data visualization module is used to display complex data and analysis results through visualization methods such as charts and graphs.

[0040] The continuous learning and optimization module is used to continuously learn and optimize based on new data and feedback to improve accuracy and effectiveness.

[0041] Furthermore, this embodiment may also include the following functional modules to further enhance the intelligence level of the management of special management places.

[0042] The educational content customization module is used to customize personalized educational content based on the educational background and interests of the interviewee.

[0043] The mental health assessment module is used to assess the mental health status of the person being interviewed through questionnaires and behavioral analysis.

[0044] The intelligent question-and-answer module is used to provide an intelligent question-and-answer interface to answer the common questions of the interviewee and reduce the workload of the interviewer.

[0045] The collection module is used to collect questions raised by the interviewee through the intelligent question-and-answer module into a database as data for subsequent intelligent analysis, which can be used, for example, for educational content customization, mental health assessment, etc.

[0046] Through these intelligent functions, the conversation education expert system in special management places can provide comprehensive support to interviewers, helping them to conduct conversation education and management more effectively, while providing more humane and personalized care to those being interviewed.

[0047] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. An intelligent conversation expert system, characterized in that: include: The pre-conversation preparation module uses a multimodal large model to mine and analyze the interviewee's background information and historical behavior data in the database, predict the interviewee's psychological state and behavioral tendencies, and generate conversation strategies and preparation suggestions based on the conversation type. The in-conversation module collects the image and voice data of the person being interviewed in real time during the conversation, analyzes the person's speech, expression and actions through a multimodal large model, predicts the person's psychological state and behavioral tendencies, and generates real-time emotional feedback information and conversation guidance suggestions based on the conversation type; The post-conversation summary module uses the learning and optimization capabilities of the multimodal large model to design a conversation content analysis module after the conversation. It summarizes the conversation effect based on the conversation type, optimizes the conversation strategy, and continuously improves the intelligence level of the system through machine learning.

2. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The knowledge base query and recommendation module is used to provide query and recommendation of professional knowledge including laws and regulations, psychological knowledge and conversation skills in the pre-conversation preparation stage or during the conversation.

3. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The case library retrieval module is used to retrieve relevant successful cases from the case library in real time during the preparation stage before the conversation or during the conversation, based on the conversation topic or the characteristics of the person being interviewed, to provide reference for the interviewer.

4. The intelligent conversation expert system according to claim 1, characterized in that: Also includes: The recording and analysis module is used to automatically record the conversation process, perform content analysis, and generate conversation reports and improvement suggestions.

5. The intelligent conversation expert system according to claim 1, characterized in that: Also includes: The data visualization module is used to display complex data and analysis results through charts and graphs.

6. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The continuous learning and optimization module is used to continuously learn and optimize based on new data and feedback to improve accuracy and effectiveness.

7. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The education content customization module is used to customize personalized education content according to the educational background and interests of the interviewee.

8. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The mental health assessment module is used to assess the mental health status of the person being interviewed through questionnaires and behavioral analysis.

9. An intelligent conversation expert system according to claim 1, characterized in that: Also includes: The intelligent question-answering module is used to provide an intelligent question-answering interface to answer the common questions of the interviewee.

10. An intelligent conversation expert system according to claim 9, characterized in that: Also includes: The collection module is used to collect the questions raised by the interviewee through the intelligent question-answering module into the database as data for subsequent intelligent analysis.

Images