Intelligent question answering system

Abstract

The invention discloses an intelligent question answering system, which relates to the technical field of intelligent question answering and comprises a question input module, an answer presentation module and a feedback collection module which are arranged on a front-end user interface, the question analysis module, the answer generation module and the knowledge base are arranged on the back-end server; the question input module is used for realizing question input; the problem analysis module can accurately analyze the meaning of the problem and identify the key point and implicit demand of the problem; the answer generation module retrieves a knowledge base to generate answers based on an analysis result of the question analysis module; the answer presentation module is used for presenting answers; the feedback collection module is used for collecting evaluation and feedback of the user on the answer and generating feedback information; and the knowledge base is used for storing domain knowledge and feedback information. According to the method, the user experience is improved, the question answering efficiency is improved, meanwhile, development of an intelligent question answering system can be promoted, and the method can adapt to diversified application scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent question answering, and in particular to an intelligent question answering system. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.

[0003] With the continuous development of artificial intelligence technology, intelligent question-answering systems have been widely used in various fields. However, there are large differences in the interaction processes of different intelligent question-answering systems, resulting in uneven user experience and difficulty in ensuring the efficiency of question-answering. The lack of unified interaction process specifications has limited the development of intelligent question-answering systems.

[0004] The difference is extremely obvious in the question input stage. Some intelligent question-answering systems only support traditional text input methods, which undoubtedly increases the difficulty of operation for users who are accustomed to using voice input. For example, in a busy work scene, the user's hands may be occupied and unable to input text. At this time, if the system does not support voice input, it will greatly affect the efficiency of the user's acquisition of information. Although some systems support voice input, the accuracy of voice recognition is not high, which causes users to need to input repeatedly for the system to correctly understand the problem, wasting a lot of time and energy. In addition, the preliminary legality verification standards for input content are also different. Some systems are too strict in verification, and some slightly special expressions may be judged as illegal, while some systems are too loose in verification and are prone to accept some unclear or even wrong input, which brings great uncertainty to the subsequent question analysis and answer generation.

[0005] Entering the question analysis stage, the natural language processing technology levels adopted by different intelligent question-answering systems vary. Some systems have major defects in semantic understanding and cannot accurately parse the true meaning of complex question statements. For example, when a user asks a question containing multiple clauses and modifiers, some systems may misunderstand the focus of the question and extract the wrong keywords. In terms of keyword extraction, the algorithms and standards of different systems are also different. Some systems focus too much on the extraction of high-frequency words and ignore some low-frequency words with specific meanings, resulting in an inability to fully grasp user needs. Some systems are easily affected by noise during the keyword extraction process, extracting some words that are irrelevant to the question, affecting the accuracy of question analysis.

[0006] The differences are also prominent in the answer generation stage. Some intelligent question-answering systems cannot cover a wide range of knowledge fields due to their incomplete knowledge base. When users ask some more professional or unpopular questions, the system cannot retrieve relevant information and cannot generate accurate answers. Some other systems have richer knowledge bases, but their intelligent algorithms are not advanced enough. They lack logic and order when generating answers. The answers are either too long and complicated for users to understand, or too short and vague to meet users' needs. In addition, there are large gaps in the accuracy and reliability of answers from different systems.

[0007] In terms of answer presentation, the singleness and inconsistency of the form also brings inconvenience to users. Some systems only present answers in text form. For some data or concepts that need to be presented intuitively, it is impossible to effectively express them through charts and other forms. Although some systems support multiple presentation forms, they are not convenient in switching and selecting, and users need to spend a lot of time to find a suitable presentation method. In addition, different systems have different layouts and typesets for answer presentation, and some are too messy, affecting the user's reading experience.

[0008] As for the feedback collection stage, the performance of different intelligent question-answering systems varies greatly. Many systems either do not set up clear feedback channels, so users have nowhere to express even if they are dissatisfied with the answers. Or they do not pay enough attention to user feedback, and even if they receive feedback, they cannot process and optimize it in a timely manner. Some systems lack effective incentive mechanisms in the feedback collection process, and users are not motivated to provide feedback, which makes it impossible for the system to continuously improve.

[0009] This lack of uniformity in the interaction process requires users to constantly adapt to new operating methods and experiences when using different intelligent question-answering systems, which increases the user's learning cost and difficulty of use. At the same time, due to the lack of unified standards, it is difficult to effectively compare and evaluate various intelligent question-answering systems, which also limits the development of the entire industry. Summary of the invention

[0010] The purpose of the present invention is to provide an intelligent question-answering system to solve the problems existing in the prior art.

[0011] The technical solution of the present invention is as follows:

[0012] An intelligent question-answering system comprises: a question input module, an answer presentation module and a feedback collection module arranged in a front-end user interface; and a question analysis module, an answer generation module and a knowledge base arranged in a back-end server;

[0013] The question input module is used to implement question input; the question analysis module can accurately parse the meaning of the question and identify the key points and implicit requirements of the question; the answer generation module retrieves the knowledge base to generate answers based on the analysis results of the question analysis module; the answer presentation module is used to present the answers; the feedback collection module is used to collect users' evaluation and feedback on the answers and generate feedback information; the knowledge base is used to store domain knowledge and feedback information.

[0014] Furthermore, the question input module provides a variety of input methods to meet the needs of different users in various scenarios;

[0015] The question input module is equipped with a strict and intelligent legality verification mechanism for verifying the legality of input questions.

[0016] Furthermore, the input method includes: text input, voice input, image input;

[0017] Perform a validity check on the input questions, including:

[0018] For text input, we not only check for grammatical errors, but also judge the rationality and clarity of the questions; for voice input, we use advanced speech recognition technology to improve the accuracy, and at the same time perform a legality check on the recognized text; for image input, we use image recognition technology to extract key information and judge whether it meets the input requirements.

[0019] Furthermore, the question analysis module adopts advanced natural language processing technology combined with deep learning algorithms to improve the accuracy of semantic understanding; for complex question statements, it can accurately parse their meanings and identify the key points and implicit requirements of the questions.

[0020] Furthermore, the question analysis module accurately extracts keywords by analyzing sentence structure, semantic relationship and context information.

[0021] Furthermore, the question analysis module has a built-in keyword extraction algorithm that comprehensively considers the frequency, importance and relevance of words in a specific field; at the same time, the question analysis module uses noise reduction technology to eliminate the interference of irrelevant words.

[0022] Furthermore, the answer generation module has an advanced intelligent algorithm, which makes the generated answers logical and coherent;

[0023] The answer generation module retrieves relevant information from the knowledge base based on the analysis results of the question analysis module and generates an answer.

[0024] Furthermore, the answer presentation module provides multiple forms of answer presentation, including text, chart, voice, and video; and automatically selects the most appropriate presentation method based on the type of question and user needs.

[0025] Furthermore, the feedback collection module has multiple feedback channels, including: online evaluation, feedback form, and email feedback;

[0026] The feedback collection module can also analyze and process user feedback in a timely manner based on the preset feedback processing mechanism, and optimize and improve the system according to the feedback results.

[0027] Furthermore, the knowledge base covers knowledge in multiple fields and is continuously updated and expanded.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. Improve user experience: Through standardized interaction processes and rich and diverse functions, users can use the intelligent question-and-answer system more easily and obtain accurate, timely and diverse answers, greatly improving user satisfaction.

[0030] 2. Improve the efficiency of question-answering: Standardized interactive processes help reduce unnecessary operations and waiting time, and improve the response speed of the intelligent question-answering system. At the same time, advanced algorithms and technologies can quickly and accurately analyze questions and generate answers, further improving the efficiency of question-answering.

[0031] 3. Promote the development of intelligent question-answering systems: Unified interactive process specifications provide common standards for different intelligent question-answering systems, which is conducive to promoting the development and innovation of the industry. Each system can continuously upgrade technology and expand functions on the basis of following the specifications to provide better services for users.

[0032] 4. Adapt to diverse application scenarios: Standardized interaction processes can ensure that intelligent question-and-answer systems in different fields have similar operation methods and user experiences, making it easier for users to switch between different scenarios. Unified interaction process specifications can enable intelligent question-and-answer systems to better adapt to the characteristics of different devices and user habits, providing a consistent service experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a block diagram of an intelligent question answering system;

[0034] Figure 2 The present invention is a flow chart of a standard interactive process specification method for intelligent question and answer. DETAILED DESCRIPTION

[0035] It should be noted that relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0036] The features and performance of the present invention are further described in detail below in conjunction with the embodiments.

[0037] Embodiment 1

[0038] See also Figure 1 , an intelligent question answering system, specifically comprising:

[0039] A question input module, an answer presentation module and a feedback collection module are provided in the front-end user interface; and a question analysis module, an answer generation module and a knowledge base are provided in the back-end server;

[0040] The question input module is used to input questions; the question analysis module can accurately analyze the meaning of the question and identify the key points and implicit requirements of the question; the answer generation module retrieves the knowledge base to generate answers based on the analysis results of the question analysis module; the answer presentation module is used to present the answers; the feedback collection module is used to collect users' evaluation and feedback on the answers and generate feedback information; the knowledge base is used to store domain knowledge and feedback information;

[0041] It should be noted that the front-end user interface: provides a friendly and intuitive interactive environment for users to enter questions and view answers. The interface design should be simple and beautiful, easy to operate. Responsive design should be used to adapt to the screen size and resolution of different devices;

[0042] Backend server: responsible for core tasks such as question analysis, answer generation and knowledge base management. It uses high-performance server architecture and advanced technology to ensure system stability and response speed.

[0043] Continuously optimize system performance, improve system reliability and scalability. Regularly update and upgrade the system, introduce new technologies and functions to meet the ever-changing needs of users.

[0044] In this embodiment, specifically, the question input module provides a variety of input methods to meet the needs of different users in various scenarios;

[0045] The question input module is equipped with a strict and intelligent legality verification mechanism for verifying the legality of input questions.

[0046] In this embodiment, specifically, the input mode includes: text input, voice input, image input;

[0047] Perform a validity check on the input questions, including:

[0048] For text input, we not only check for grammatical errors, but also judge the rationality and clarity of the questions; for voice input, we use advanced speech recognition technology to improve the accuracy, and at the same time perform a legality check on the recognized text; for image input, we use image recognition technology to extract key information and judge whether it meets the input requirements.

[0049] In this embodiment, specifically, the question analysis module adopts advanced natural language processing technology combined with deep learning algorithms to improve the accuracy of semantic understanding; for complex question statements, it can accurately parse their meanings and identify the key points and implicit requirements of the questions.

[0050] In this embodiment, specifically, the question analysis module accurately extracts keywords by analyzing sentence structure, semantic relationship and context information.

[0051] In this embodiment, specifically, the question analysis module has a built-in keyword extraction algorithm that comprehensively considers the frequency, importance and relevance of words in a specific field; at the same time, the question analysis module uses noise reduction technology to eliminate the interference of irrelevant words.

[0052] In this embodiment, specifically, the answer generation module has an advanced intelligent algorithm, so that the generated answer is logical and coherent;

[0053] The answer generation module retrieves relevant information from the knowledge base based on the analysis results of the question analysis module and generates an answer.

[0054] In this embodiment, specifically, the answer presentation module provides multiple forms of answer presentation, including: text, chart, voice, video; according to the type of question and the needs of the user, the most appropriate presentation method is automatically selected.

[0055] In this embodiment, specifically, the feedback collection module has multiple feedback channels, including: online evaluation, feedback form, and email feedback;

[0056] The feedback collection module can also analyze and process user feedback in a timely manner based on the preset feedback processing mechanism, and optimize and improve the system according to the feedback results.

[0057] In this embodiment, specifically, the knowledge base covers knowledge in multiple fields and is continuously updated and expanded.

[0058] In this embodiment, it should also be noted that the above-mentioned intelligent question and answer system is designed and developed based on an intelligent question and answer standard interaction process specification method.

[0059] In this embodiment, please refer to Figure 2 , a method for standardizing an intelligent question-answering standard interaction process, comprising:

[0060] Step S1: Design of interaction process of intelligent question-answering system;

[0061] The interactive process of the intelligent question-answering system is divided into five main stages: question input, question analysis, answer generation, answer presentation, and feedback collection;

[0062] Among them, the problem input stage:

[0063] Provide a variety of input methods, including text input, voice input, image input, etc., to meet the needs of different users in various scenarios. For example, users can use voice input to ask questions while driving, and use image input to ask questions related to pictures when viewing picture materials;

[0064] Specifically, for text input, advanced text input box technology is used to support automatic error correction, intelligent prompts and other functions. For example, when a user makes a spelling error during input, the system automatically gives the correct spelling suggestion. At the same time, an input length limit is set to avoid the problem of users inputting too long or too short.

[0065] For voice input, a high-quality voice recognition engine is integrated, which can adapt to different accents and speaking speeds. When the user inputs voice, the system displays the recognition results in real time for user confirmation. In addition, the input voice is processed for noise reduction to improve recognition accuracy;

[0066] For image input, we use image recognition technology to analyze the images uploaded by users and extract key information. For example, for images containing text, we use optical character recognition technology to extract the text and analyze it as part of the question.

[0067] Establish a strict and intelligent legality verification mechanism. For text input, not only grammatical errors should be checked, but also the rationality and clarity of the questions should be judged. For voice input, advanced voice recognition technology is used to improve the accuracy, and the legality of the recognized text is verified. For image input, key information is extracted through image recognition technology, and it is judged whether it meets the input requirements;

[0068] Problem analysis phase:

[0069] Use lexical analysis, syntactic analysis, and semantic analysis techniques in natural language processing to gain a deep understanding of the input questions. Extract keywords, topics, and intent from the questions to provide an accurate basis for subsequent answer generation. Use advanced natural language processing technology combined with deep learning algorithms, such as large language models, to improve the accuracy of semantic understanding. For complex question statements, it can accurately parse their meanings and identify the key points and implicit requirements of the questions. For example, by analyzing sentence structure, semantic relationships, and contextual information, it can accurately extract keywords;

[0070] Establish a problem classification system to classify different types of problems so that different processing strategies can be adopted. For example, for factual questions, the answers can be directly retrieved from the knowledge base; for reasoning questions, logical reasoning algorithms need to be used to solve them;

[0071] Optimize the keyword extraction algorithm, comprehensively consider the frequency, importance and relevance of words in specific fields. Not only pay attention to high-frequency words, but also do not miss low-frequency but key words. At the same time, use noise reduction technology to eliminate the interference of irrelevant words to ensure that the extracted keywords can accurately reflect user needs;

[0072] Answer generation phase:

[0073] Build a rich and comprehensive knowledge base, including: structured data (such as table data in the database) and unstructured data (such as text files, web page content, etc.); cover knowledge in multiple fields, and continuously update and expand. Ensure that when users ask professional or unpopular questions, the system can also retrieve relevant information from the knowledge base; use knowledge graph technology to associate and integrate the knowledge in the knowledge base so as to retrieve relevant information more quickly;

[0074] According to the results of question analysis, relevant knowledge fragments are retrieved from the knowledge base, and intelligent algorithms (such as the GPT series model) are used to generate and optimize answers. Ensure that the generated answers are accurate, concise, and easy to understand;

[0075] Develop advanced intelligent algorithms to make the generated answers logical and organized. The answers should not be too long and complicated, nor too short and vague, and should be able to answer user questions accurately and concisely. In the process of generating answers, strict verification and screening are carried out to ensure the accuracy and reliability of the answers;

[0076] Answer presentation stage:

[0077] Provide multiple forms of answer presentation, including text, charts, voice, video, etc. Automatically select the most appropriate presentation method based on the type of question and user needs; for example, for simple questions, the answer can be presented directly in text form; for data-related questions, it can be presented intuitively with charts; for complex questions, it can be presented in the form of charts, images or animated videos to enhance the intuitiveness and readability of the answer;

[0078] The text is presented in clear and readable fonts and layouts. Important content can be bolded, changed in color, etc. to attract the user's attention. At the same time, hyperlinks and references are supported to facilitate users to further understand related content;

[0079] Chart presentation Choose appropriate chart format according to the type of question, such as bar chart, line chart, pie chart, etc. Use data visualization technology to present complex data to users in an intuitive form;

[0080] The voice presentation uses natural and smooth speech synthesis technology and supports multiple languages ​​and timbre options. Users can set the speech speed and volume according to their preferences;

[0081] Optimize the layout and typesetting of the presentation interface to make the answers easy to read and understand. Set up a convenient presentation mode switch button to facilitate users to switch between different forms; at the same time, provide relevant links and reference materials to help users further understand the background and relevant knowledge of the questions.

[0082] Feedback collection phase:

[0083] Establish multiple feedback channels, such as online evaluation, feedback forms, email feedback, etc., to ensure that users can easily evaluate and provide feedback on answers; for example, after the answer is presented, provide a feedback button or evaluation area to encourage users to evaluate and provide feedback on the answer. Feedback content can include accuracy, completeness, and understandability of the answer;

[0084] We attach great importance to user feedback and have established a special feedback processing mechanism. We analyze and process user feedback in a timely manner, and optimize and improve the system based on the feedback results; we promptly correct and optimize wrong answers and add them to the knowledge base; we promptly add new questions raised by users to the knowledge base to continuously improve the intelligent question-answering system;

[0085] Step S2: Formulate standardization specifications for the intelligent question-answering interaction process.

[0086] In this embodiment, specifically, step S2 includes:

[0087] Step S21: Formulate detailed and strict standardization specifications for each stage of the interaction process;

[0088] Specifically, question input format requirements: clarify the format standards of different input methods, such as character limits for text input, duration and sound quality requirements for voice input, resolution and format requirements for image input, etc.;

[0089] Question analysis algorithm standards: specify the specific algorithms and technical indicators for semantic understanding and keyword extraction to ensure consistency and accuracy in question analysis across different systems;

[0090] Answer generation rules: formulate the principles and standards for answer generation, including requirements for answer accuracy, completeness, logic, length and language style of answers, etc.;

[0091] Answer presentation format: Determine the specific format requirements for various presentation methods, such as the font, size, and color specifications of the text, the type and style standards of the charts, the speed and timbre requirements of the voice, etc.

[0092] Step S22: Ensure that different intelligent question-answering systems can achieve consistency and efficiency in the interactive process under the premise of following these specifications. Through the formulation of standardized specifications, different systems can be effectively compared and evaluated, promoting the healthy development of the entire industry.

[0093] The above-mentioned embodiments only express the specific implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the protection scope of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the technical solution concept of the present application, and these all belong to the protection scope of the present application.

[0094] This background section is provided to generally present the context of the invention, and the work of the presently named inventors, the work to the extent described in this background section, and aspects of this section that did not constitute prior art at the time of application are neither explicitly nor implicitly admitted to be prior art to the present invention.

Claims

1. An intelligent question-answering system, characterized in that: include: A question input module, an answer presentation module and a feedback collection module are set up in the front-end user interface; and a question analysis module, an answer generation module and a knowledge base arranged on a back-end server; The question input module is used to input questions; the question analysis module can accurately analyze the meaning of the question and identify the key points and implicit requirements of the question; the answer generation module searches the knowledge base to generate answers based on the analysis results of the question analysis module; the answer presentation module is used to present answers; the feedback collection module is used to collect users' evaluation and feedback on the answers and generate feedback information; The knowledge base is used to store domain knowledge and feedback information.

2. An intelligent question-answering system according to claim 1, characterized in that: The question input module provides a variety of input methods to meet the needs of different users in various scenarios; The question input module is equipped with a strict and intelligent legality verification mechanism for verifying the legality of input questions.

3. An intelligent question-answering system according to claim 2, characterized in that: Input methods, including text input, voice input, and image input; Perform a validity check on the input questions, including: For text input, we not only check for grammatical errors, but also judge the rationality and clarity of the questions; for voice input, we use advanced speech recognition technology to improve the accuracy, and at the same time perform a legality check on the recognized text; for image input, we use image recognition technology to extract key information and judge whether it meets the input requirements.

4. An intelligent question-answering system according to claim 3, characterized in that: The question analysis module uses advanced natural language processing technology combined with deep learning algorithms to improve the accuracy of semantic understanding; for complex question statements, it can accurately parse their meanings and identify the key points and implicit requirements of the questions.

5. An intelligent question-answering system according to claim 4, characterized in that: The question analysis module accurately extracts keywords by analyzing sentence structure, semantic relationships and context information.

6. An intelligent question-answering system according to claim 5, characterized in that: The question analysis module has a built-in keyword extraction algorithm that comprehensively considers the frequency, importance and relevance of words in a specific field; at the same time, the question analysis module uses noise reduction technology to eliminate the interference of irrelevant words.

7. An intelligent question-answering system according to claim 6, characterized in that: The answer generation module has an advanced intelligent algorithm to make the generated answers logical and organized; The answer generation module retrieves relevant information from the knowledge base based on the analysis results of the question analysis module and generates an answer.

8. An intelligent question-answering system according to claim 7, characterized in that: The answer presentation module provides multiple forms of answer presentation, including text, chart, voice, and video; it automatically selects the most appropriate presentation method based on the type of question and user needs.

9. An intelligent question-answering system according to claim 8, characterized in that: The feedback collection module has multiple feedback channels, including: online evaluation, feedback form, and email feedback; The feedback collection module can also analyze and process user feedback in a timely manner based on the preset feedback processing mechanism, and optimize and improve the system according to the feedback results.

10. An intelligent question-answering system according to claim 9, characterized in that: The knowledge base covers knowledge in multiple fields and is constantly updated and expanded.

Images