Research data processing method, system and equipment based on AI analysis and storage medium

Abstract

The invention discloses a survey data processing method, system and device based on AI analysis and a storage medium. According to the method, investigation mode information can be obtained, and a data analysis source is determined according to the investigation mode information. And creating an interactive dialogue to obtain question data input by the user. And inputting the question data into the AI analysis model, so that the AI analysis model extracts survey project data from the question data, and obtains key information from the data analysis source according to the survey project data to output answer data and associated data. And displaying the answer data and the associated data. According to the method, the investigation data can be processed based on artificial intelligence, automation and data analysis technologies, and the processing efficiency and accuracy of the investigation data are effectively improved. And key information is obtained from different data analysis sources according to different investigation modes, so that a personalized data analysis method can be provided, and the pertinence and effectiveness of investigation and research are enhanced.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a survey data processing method, system, device and storage medium based on AI analysis. Background Art

[0002] Research is the abbreviation of investigation and research. It is to obtain research data in a specific field or topic by reading and analyzing a certain amount of literature. Research data can be used to understand a specific field or problem, and assist in decision support, problem identification, market analysis, product development, policy formulation, academic research, risk and effect assessment, resource allocation, etc.

[0003] In the process of conducting investigation and research, research tools are needed. Research tools refer to tools used to collect, analyze and present data. Research tools include data collection tools, data analysis tools, data storage and management tools, etc. Research tools can facilitate research work, but they also have some disadvantages. For example, data collection tools have the problem of low data quality. Data analysis tools require human participation, which not only consumes a lot of time and energy, but also leads to deviations in analysis results due to human factors. Data storage and management tools have problems such as low data security and low data query efficiency.

[0004] In view of the problems existing in the survey tools, the literature can also be analyzed and sorted based on artificial intelligence (AI) data analysis technology. When analyzing and sorting the literature based on artificial intelligence data analysis technology, the large model can be pre-trained according to the set analysis objectives. That is, the large model of the Transformer, BERT and other structures is trained using the training data set, and the model is adjusted through cross-validation, hyperparameter optimization and other methods to improve the performance and generalization ability of the model. After the large model is trained, the large model reads the relevant data by inputting data such as the purpose of the survey and the survey object, and outputs the corresponding survey results after sorting. However, when dealing with complex problems, artificial intelligence data analysis technology will be affected by the quality, completeness and accuracy of the data, resulting in low processing efficiency and the risk of data loss or damage when processing large amounts of data. In addition, it is difficult to cope with complex and changeable real-world situations, and lacks pertinence and effectiveness. Summary of the invention

[0005] In view of this, the embodiments of the present application provide a survey data processing method, system, device and storage medium based on AI analysis to provide personalized data support and analysis methods to enhance the pertinence and effectiveness of the survey and research.

[0006] According to one aspect of the present application, a survey data processing method based on AI analysis is provided, the method comprising:

[0007] In response to a survey instruction input by a user, obtaining survey method information, wherein the survey method information is used to represent the survey method specified by the user;

[0008] Determine a data analysis source according to the survey method information, wherein the data analysis source has an association relationship with the survey method corresponding to the survey method information;

[0009] Creating an interactive dialogue, and acquiring question data input by a user based on the interactive dialogue;

[0010] The question data is input into an AI analysis model to obtain answer data and associated data output by the AI ​​analysis model, wherein the AI ​​analysis model is a question-answering model trained using research literature training data; the AI ​​analysis model is used to extract research project data from the question data, and to obtain key information in the data analysis source based on the research project data;

[0011] The answer data and the associated data are displayed based on the interactive dialog.

[0012] Optionally, the method further includes:

[0013] Acquire research document training data, wherein the research document training data includes question-answer pairs generated based on the research documents; the question-answer pairs include a question text and an answer text;

[0014] Inputting the question text in the research literature training data into the trained model to obtain the answer result data output by the trained model, wherein the input of the trained model is the question text of natural language type; the output of the trained model is the answer text of natural language text type;

[0015] Calculate training loss based on the answer text and the answer result data;

[0016] If the training loss is greater than a loss threshold, adjusting a model parameter of the trained model according to the training loss;

[0017] If the training loss is less than or equal to the loss threshold, the model parameters of the trained model are output to obtain the AI ​​analysis model.

[0018] Optionally, the method further includes:

[0019] Obtain literature data;

[0020] Performing text cleaning and word segmentation processing on the document data to generate a word set;

[0021] Converting the words in the word set into word vectors of fixed dimension based on a word embedding algorithm;

[0022] Acquire context information, and perform word vector enhancement on the word vector based on the context information to obtain a word vector sequence;

[0023] Extracting semantic information from the word vector sequence using a neural network to generate the document data feature vector, wherein the feature vector includes a word granularity feature vector, a topic granularity feature vector, and a character granularity feature vector;

[0024] Feature fusion and pooling processing are performed on the feature vector to generate a text vector representation of a fixed length.

[0025] Optionally, obtain information about the survey method, including:

[0026] Displaying a research interface, the research interface includes an input control and a mode selection control, the mode selection control includes a first control, a second control and a third control, the research mode represented by the first control is a full-library research; the research mode represented by the second control is a special research; the research mode represented by the third control is a single article study;

[0027] Obtaining an operation event input by a user based on the survey interface;

[0028] If the operation event is based on the input control input, obtaining user input information, and extracting survey method information from the input information;

[0029] If the operation event is based on the mode selection control input, the survey mode information is generated based on the mode selection control selected by the operation event.

[0030] Optionally, the method further includes:

[0031] Detect the current user login status;

[0032] If the login status is logged in, executing the step of displaying the survey interface;

[0033] If the login status is not logged in, the login interface is displayed, and the user information input by the user based on the login interface is verified. After the user information login verification is passed, the step of displaying the survey interface is executed.

[0034] Optionally, determining a data analysis source according to the survey method information includes:

[0035] parsing the survey method specified by the user from the survey method information;

[0036] If the survey method is a full-database survey, determining that the data analysis source is a full-text literature database;

[0037] If the research method is a special research, determining that the data analysis source is a special document vector library, the special document vector library is used to store knowledge information in the research documents in the form of vectors;

[0038] If the research method is single article study, the data analysis source is determined to be an input document, and the input document is a document file input by a user or a document link in a document full-text database specified by a user.

[0039] Optionally, the question data is input into an AI analysis model to obtain answer data and associated data output by the AI ​​analysis model, including:

[0040] Calling the AI ​​analysis model, wherein the AI ​​analysis model includes a natural language processing module and an information extraction module;

[0041] Converting the question data into semantic information through the natural language processing module;

[0042] Determine the survey project data according to the semantic information;

[0043] Extracting key information from the data analysis source by the information extraction module, wherein the key information is information in the data analysis source associated with the survey project data;

[0044] The answer data is generated according to the extracted key information and the question data.

[0045] Optionally, the method further includes:

[0046] The question data is segmented to obtain index data, wherein the index data includes at least first-category keywords and second-category keywords; the first-category keywords are used to represent the research project data; and the second-category keywords are used to represent the analysis intention;

[0047] generating a query vector according to the index data;

[0048] Comparing the query vector with feature vectors of document data stored in a vector database to obtain relevant knowledge information, wherein the vector database is a database associated with the AI ​​analysis model and used to store feature vectors associated with document data;

[0049] The answer data is generated based on the relevant knowledge information.

[0050] Optionally, generating the answer data based on the relevant knowledge information includes:

[0051] Extracting key information from the relevant knowledge information, wherein the key information at least includes literature information corresponding to the research project data and analysis result information corresponding to the analysis intention;

[0052] Setting the information points and logical order of the key information;

[0053] Based on the key information after setting the information points and the logical order, an answer text is generated according to an answer sentence template, and the answer sentence template includes one or more combinations of a preset sentence template, a preset rule template, and a machine learning algorithm template.

[0054] Optionally, the method further includes:

[0055] Record the source document information of the key information;

[0056] Extracting keywords from the key information, and matching synonyms according to the keywords;

[0057] Performing a literature search based on the keywords and / or the synonyms to obtain related literature information;

[0058] The associated data is generated according to the extracted source document information and the associated document information.

[0059] Optionally, if the research method is a special research, the method further includes:

[0060] Get logged-in user information;

[0061] Create a new subject document vector library according to the user information;

[0062] Acquire special document information, which includes search documents, uploaded documents and knowledge base. The search documents are documents obtained by performing document search based on the special index input by the user; the uploaded documents are documents uploaded by the user through the research interface; the knowledge base is an information collection composed of knowledge information;

[0063] Generate a document vector according to the subject document information, wherein the document vector includes multi-dimensional knowledge information extracted from the document information and traceability information associated with the knowledge information;

[0064] The document vector is stored in the document vector library.

[0065] According to another aspect of the present application, a survey data processing system based on AI analysis is provided, the system comprising:

[0066] A survey method acquisition module is configured to acquire survey method information in response to a survey instruction input by a user, wherein the survey method information is used to represent the survey method specified by the user;

[0067] An analysis source module is configured to determine a data analysis source according to the survey method information, wherein the data analysis source is associated with the survey method corresponding to the survey method information;

[0068] A dialogue module is configured to create an interactive dialogue and obtain question data input by the user based on the interactive dialogue;

[0069] The AI ​​analysis module is configured to input the question data into an AI analysis model to obtain answer data and related data output by the AI ​​analysis model, wherein the AI ​​analysis model is a question-answering model trained using research literature training data; the AI ​​analysis model is used to extract research project data from the question data, and to obtain key information from the data analysis source based on the research project data;

[0070] The display module is configured to display the answer data and the associated data based on the interactive dialogue.

[0071] According to another aspect of the present application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned survey data processing method based on AI analysis when executing the program.

[0072] According to another aspect of the present application, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned survey data processing method based on AI analysis is implemented.

[0073] By means of the above technical scheme, the embodiment of the present application provides a survey data processing method, system, device and storage medium based on AI analysis. The method can obtain survey method information and determine the data analysis source according to the survey method information. Then create an interactive dialogue to obtain the question data input by the user. The question data is input into the AI ​​analysis model, so that the AI ​​analysis model extracts the survey project data from the question data, and obtains key information from the data analysis source according to the survey project data to output the answer data and related data. Then display the answer data and related data. The method can process the survey data based on artificial intelligence, automation and data analysis technology, and effectively improve the processing efficiency and accuracy of the survey data. And according to different survey methods, key information is obtained from different data analysis sources, which can provide personalized data analysis methods and enhance the pertinence and effectiveness of the investigation and research.

[0074] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0076] Figure 1 A schematic diagram of an application scenario of a survey data processing method based on AI analysis provided in an embodiment of the present application is shown;

[0077] Figure 2 A schematic diagram of a process of a survey data processing method based on AI analysis provided in an embodiment of the present application is shown;

[0078] Figure 3 A schematic diagram of a data processing flow of a survey data processing method provided in an embodiment of the present application is shown;

[0079] Figure 4 A schematic diagram of a survey interface provided in an embodiment of the present application is shown;

[0080] Figure 5 A schematic diagram of a process for determining a data analysis source provided by an embodiment of the present application is shown;

[0081] Figure 6 A schematic diagram of the interactive dialogue display effect provided by an embodiment of the present application is shown;

[0082] Figure 7 A schematic diagram of a process for generating answer data provided in an embodiment of the present application is shown;

[0083] Figure 8 A schematic diagram of a flow chart of information generation of a survey method provided in an embodiment of the present application is shown;

[0084] Fig. 9 A schematic diagram of a login status detection process provided in an embodiment of the present application is shown;

[0085] Fig.10 A schematic diagram of the structure of a survey data processing system based on AI analysis provided in an embodiment of the present application is shown;

[0086] Fig.11 A schematic diagram of the structure of a computer device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0087] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0088] In the embodiments of the present application, the survey data is a general term for various data generated during the investigation and research of literature. The survey data has a broad meaning and may include but is not limited to the data to be processed, the data in process, and the processing result data used or generated during the investigation and research.

[0089] Among them, the data to be processed is the data directly derived from the literature. For example, the content fragments of the literature, the statistical data in the literature, etc. The data being processed is the intermediate data calculated by the research tool during the data processing process during the research and analysis process. For example, knowledge information, keywords, document tags, intermediate values, etc. The processed result data is the data obtained after data analysis and can be used to present to the user, such as literature review, literature framework analysis results, abbreviated text, continued text, expanded text, and polished text, etc.

[0090] The document data described in the embodiments of the present application refers to a collection of various text materials used for investigation and research. The document data may include at least one source file of the document data. Depending on the field of investigation and research, the document data may specifically include different contents. In certain embodiments, the document data may include professional documents, laws and regulations, policy documents, and Internet documents, etc.

[0091] Document data can come from different data sources, such as local databases, online public databases, third-party databases, user input, etc. The document presentation and data size under different data sources are somewhat different. For example, for the C-net professional document database, it can include newspapers, journals, conference papers, yearbooks and other professional document data. The online public database can include web pages, shared files, etc.

[0092] Literature data can be processed by research tools and converted into research data as the processing progresses. Among them, research tools include data collection tools, data analysis tools, data storage and management tools, etc. Users can call research tools according to specific needs and perform corresponding data processing through research tools to obtain research data processing results. For example, before performing data analysis processing, you can call data collection tools to obtain literature data from the selected data source. After obtaining the literature data, call data analysis tools to analyze and process the literature data to obtain processing results. After obtaining the processing results, call data storage and management tools to store and display the processing results.

[0093] The survey tool can be operated based on a specific data processing or analysis method to collect, transform and process literature data and survey data. In some embodiments, the survey tool can be constructed based on an AI analysis method. The AI ​​analysis method is to process survey data through an artificial intelligence algorithm to achieve functions such as statistics, analysis, and prediction. The AI ​​analysis method may include one or more combinations of algorithms such as machine learning, deep learning, natural language processing (NLP), evolutionary algorithms, knowledge-based and reasoning methods, decision-making algorithms, search methods and optimization theories, data visualization, data preprocessing, and data collection based on AI algorithms.

[0094] Different algorithms can be used in different research processes and at different stages. Along with the execution of AI analysis methods, research tools also use automation technology and data analysis technology to process research data at multiple levels to obtain richer processing results.

[0095] Because some AI data analysis processes have certain deficiencies in data quality, completeness and accuracy when dealing with complex problems. As a result, when processing large amounts of data, there is a risk of low processing efficiency, data loss or damage. In addition, due to the difficulty in coping with complex and changeable real-world situations, it lacks pertinence and effectiveness. Therefore, an embodiment of the present application provides a survey data processing method based on AI analysis. The survey data processing method can be applied to a survey data processing system based on AI analysis.

[0096] The survey data processing system is a software and hardware platform that performs data processing based on literature data or user input data and displays the data processing results. The survey data processing system can rely on computers, mobile terminals, servers, industrial hosts, smart wearable devices, etc. to provide users with a human-computer interaction interface, and in response to user input, perform operations such as collection, processing, transmission, and storage on relevant data based on the provided human-computer interaction interface to meet specific user needs.

[0097] In a feasible implementation, the survey data processing system can be run on a single device, that is, the user can realize all the functions of the survey data processing system through a single client device. For example, the survey data processing system is an application installed on a personal computer. The survey data processing system application has an application program with a survey data processing algorithm built in. When performing survey data processing, the computer can obtain document data and / or user input data, and call the AI ​​analysis model to perform analysis and processing on the document data to obtain the corresponding analysis and processing result data. Then, based on the display program provided by the application, the analysis and processing result data is displayed.

[0098] In another feasible implementation, the survey data processing system can be operated by multiple devices, that is, the coordination between multiple devices is required to realize all the functions of the survey data processing system. Figure 1 As shown, the survey data processing system includes a server and multiple clients connected to the server and multiple data storage units, and the multiple data storage units can be used to store different types of document data, user data, and survey data. Multiple clients can be operated by different users respectively, and the interactive data generated by the operation can be sent to the server through the network to request the server to collect, process, evaluate, distribute, and store the processing result data of the document data. After receiving the processing result data distributed by the server, the client displays the processing result data based on the display program provided by the application.

[0099] It should be noted that in the embodiments of the present application, a computer or other client is used as an example to illustrate the execution subject of the survey data processing method. Obviously, the survey data processing method can be implemented by running relevant applications on a computer or other client, or by running relevant applications on a server, or by cooperating with a client to run relevant applications. Therefore, from the description process of the execution subject, the steps corresponding to the survey data processing method can be executed by the client, or by the server, or some steps can be executed by the client and some steps can be executed by the server.

[0100] like Figure 2 , Figure 3 As shown, this embodiment provides a survey data processing method based on AI analysis, comprising the following steps:

[0101] S101. In response to a survey instruction input by a user, obtain survey method information.

[0102] When executing the survey data processing method, the client or server may provide a user interaction interface according to the application corresponding to the survey data processing method, and obtain multiple interaction instructions input by the user based on the user interaction interface. Among them, the instructions used to control the survey data processing system to perform survey data processing are called survey instructions.

[0103] Depending on the type of user interaction interface provided by the survey data processing system, the user can input the survey instruction in different ways. For example, the survey data processing system can provide a user interaction interface for survey, i.e., a survey interface. Figure 4 As shown, the survey interface may include a survey instruction input control. When the user performs an interactive action based on the survey instruction input control, an operation event containing information of the survey instruction input control is generated. At this time, the survey data processing system can obtain the survey instruction.

[0104] The survey instruction can also be directly input through natural language text. That is, in some embodiments, the survey interface provided by the survey data processing system may include a text input control. When the user inputs text data containing the survey research intention based on the text input control, the survey data processing system can generate a survey instruction by semantically recognizing the input data, and when the semantic recognition finds that the current input data has the survey operation intention.

[0105] According to the specific interaction methods supported by the client or server of the research data processing system, the research instructions can also be input in other ways. For example, when the client supports voice interaction, the user can input the research instructions by inputting voice interaction instructions such as "Please analyze the research status of project B in field A", that is, the research instructions can be input through voice interaction.

[0106] After obtaining the survey instruction input by the user, the survey data processing system can respond to the survey instruction and obtain survey method information. The survey method information is used to characterize the survey method specified by the user. In some embodiments, the survey method may include a full library survey, a special topic survey, and a single article study. The survey instruction input by the user may include survey method information, or the survey method may be specified when the survey instruction is input.

[0107] The research method information can be represented by a specific identification code. For example, the research method information corresponding to the whole database research contains the field "Type of analysis = 00"; the research method information corresponding to the special research contains the field "Type of analysis = 01"; the research method information corresponding to the single article study contains the field "Type of analysis = 10". After obtaining the research instruction, the client can read the field used to represent the research method in the research instruction to obtain the research method information.

[0108] S102. Determine a data analysis source according to the survey method information.

[0109] After obtaining the survey method information, the client or server can determine the data analysis source according to the survey method information. The data analysis source refers to the basic database when performing data analysis. The data analysis source is associated with the survey method corresponding to the survey method information. That is, the data analysis source corresponding to the full-database survey is the document full-text database. The database range of the document full-text database covers all the data that can be obtained by the entire system. For example, the data analysis source corresponding to the full-database survey may include the c-net document full-text database and the policy and regulations database.

[0110] The data analysis source corresponding to the special research is the special document vector library, which is used to store the knowledge information in the research documents in the form of vectors. The document vector can include multi-dimensional knowledge information extracted from the document data and the traceability information associated with the knowledge information. By constructing a special document vector library, the knowledge information in the document is stored in the form of vectors, which can realize fast question-answering and accurate query of specific topics, and support knowledge traceability and related document recommendations, thereby improving targeted research capabilities and meeting users' needs for subject area knowledge.

[0111] In some embodiments, if the research method is a special research, the method can also obtain the login user information, and create a special document vector library according to the user information, then obtain the special document information, generate a document vector according to the special document information, and store the document vector in the document vector library.

[0112] For example, after the user clicks the "Special Research" control in the research interface, the user can specify the research method as a special research. At this time, the research data management system can obtain the currently logged-in user information and query the special document vector library based on the logged-in user information. If the special document vector library is found, the special document vector library can be called for subsequent information extraction and generation of analysis results. If the special document vector library is not found, a new special document vector library can be created based on the logged-in user information. After the special document vector library is created, the document data is obtained based on the user's further interactive operations, and the document data is converted into document vectors and stored in the special document vector library.

[0113] The subject document information includes search documents, uploaded documents and knowledge base. The search documents are documents obtained by performing document search based on the subject index input by the user. Depending on the type of interactive operation performed by the user, the survey data management system can obtain subject document information in different ways. For example, after creating a new subject document vector library, the survey data management system can obtain the search text entered by the user for performing document search, including keywords, search terms, etc. Then, the document search is performed based on the input search text, and the document data obtained by the search is recorded to generate a document vector.

[0114] Uploaded documents are document data uploaded by users through the research interface. For example, after creating a new thematic document vector library, the research data management system can obtain the document data uploaded by users, including document source files, document fragment texts, etc. Uploaded documents can also be used to generate document vectors, thereby building a thematic document vector library.

[0115] The knowledge base is an information collection composed of knowledge information. That is, the research data management system can directly obtain knowledge information based on user input or a third-party document vector library, and then transform the acquired knowledge information to generate a document vector that meets the vector form of the subject document vector library.

[0116] The data analysis source corresponding to a single study is the input document. The specific form of the input document can be a single document file or a document collection consisting of multiple documents. In some embodiments, the input document can also include the traceability information of the specific document. The traceability information is used to locate the specific document. For example, the traceability information can be the title of a document, the unique code of a document, the storage address of a document in a full-text database or a special document vector library, etc.

[0117] like Figure 5As shown, in some embodiments, in order to determine the data analysis source, the client or server can parse the user-specified research method from the research method information. If the research method is a full-library research, the data analysis source is determined to be a full-text database of documents. If the research method is a special research, the data analysis source is determined to be a special document vector library. If the research method is a single article study, the data analysis source is determined to be an input document, and the input document is a document file input by the user or a document link in a full-text database specified by the user.

[0118] S103: Create an interactive dialogue, and obtain question data input by the user based on the interactive dialogue.

[0119] After determining the data analysis source, the survey data processing system can create an interactive dialogue, where the interactive dialogue is the communication channel between the artificial intelligence system and the user. In the interactive dialogue, the user can ask questions, that is, input question data, and the AI ​​analysis model will give corresponding answer data based on the specific content of the question data. Then, in the form of one question and one answer, the question data and answer data are displayed in the interactive dialogue.

[0120] Interactive dialogue is interactive, that is, the dialogue is two-way, and users can communicate with the AI ​​analysis model in real time. The AI ​​analysis model can understand the question data entered by the user and give feedback. Interactive dialogue can display the question data and answer data in the form of natural language. And the data in multiple rounds of dialogue can be recorded so that the AI ​​analysis model can understand the context of the dialogue based on the recorded content of multiple rounds of dialogue, so as to maintain coherence in multiple rounds of dialogue. In addition, interactive dialogue can understand the user's historical interactions and preferences by recording the content of the dialogue, so that the AI ​​analysis model can provide personalized answers.

[0121] The created interactive dialogue can include text input controls and answer display controls in the survey interface. Among them, the text input control can be a text input box for users to input question data. The answer display control can be a display view in the survey interface, which is used to display answer data. For example, when the user clicks the "Full Library Survey" control, the text input box can be displayed in the survey interface first. After the user enters the question data "Please analyze the research status of project B in field A" based on the text input box, the survey data processing system can obtain the question data entered by the user in the text input box for subsequent AI analysis. At the same time, switch to the interactive dialogue interface, such as Figure 6 As shown, the question data is displayed in the upper area of ​​the text input box, and the answer display view is displayed below the question data display area.

[0122] S104: Input the question data into an AI analysis model to obtain answer data and related data output by the AI ​​analysis model.

[0123] After obtaining the question data input by the user based on the interactive dialogue, the survey data processing system can perform literature data analysis based on the input question data and the determined data analysis source. Therefore, the survey data processing system can call the AI ​​analysis model and input the question data into the AI ​​analysis model. The AI ​​analysis model is a question-answering model trained using the survey literature training data.

[0124] The AI ​​analysis model is used to extract survey project data from the question data, and to obtain key information from the data analysis source based on the survey project data. As a question-answering model, the AI ​​analysis model can understand the user's questions through natural language processing technology (NLP) and find the most appropriate answers from the given information. In the process of performing AI analysis, the AI ​​analysis model can first understand the questions raised by the user, including identifying the intent of the question, key information, etc. Then, based on the determined data analysis source, retrieve information related to the question from the knowledge base or data set corresponding to the data analysis source. Thus, based on the retrieved information, the model generates one or more possible answers. The generated answers are then optimized by deduplication, aggregation, etc. to provide more accurate and comprehensive answers.

[0125] In some embodiments, the AI ​​analysis model may include basic network structures such as an encoder, an attention mechanism, a decoder, and an output layer. Among them, the encoder is responsible for converting the input question into an internal representation that the model can understand. The attention mechanism is used to help the model focus on the key parts of the question to better understand the problem. The decoder is responsible for converting the internal representation into an answer. The output layer is used to generate the final answer. For example, the AI ​​analysis model can be a question-answering model based on TensorFlow, so the AI ​​analysis model may include an encoder and a decoder, where the encoder processes the question and the decoder generates the answer based on the encoded question.

[0126] Before using the AI ​​analysis model to generate answer data, the AI ​​analysis model needs to be trained. In some embodiments, the survey data processing system can first obtain survey document training data. The survey document training data includes question-answer pairs generated based on the survey document, and the question-answer pairs include question text and answer text.

[0127] Then, the question text in the research literature training data is input into the trained model to obtain the answer result data output by the trained model. In order to perform research data processing, the trained model is an initial model or an unconverged model with the same structure as the AI ​​model. Therefore, the input of the trained model is a question text of natural language type, and the output of the trained model is an answer text of natural language text type.

[0128] After obtaining the answer result data output by the trained model, the training loss is calculated based on the answer text and the answer result data. The training loss is compared with the preset loss threshold. If the training loss is greater than the loss threshold, the model parameters of the trained model are adjusted according to the training loss, and iterative training is continued. If the training loss is less than or equal to the loss threshold, the model parameters of the trained model are output to obtain the AI ​​analysis model.

[0129] For example, the survey data processing system can first collect and preprocess the question-answer data, including questions and corresponding answers. Then perform model initialization, that is, randomly initialize the model parameters. And define the loss function, that is, define a loss function to measure the difference between the answer predicted by the model and the true answer, such as the cross entropy loss function. Then select an optimizer, such as Adam, to update the model parameters, and perform a training cycle with minimizing the loss function as the iterative goal, so as to iterate the training data multiple times and continuously update the model parameters until the model performance reaches a satisfactory level.

[0130] During the training process of the AI ​​analysis model, the trained AI analysis model can also be evaluated and adjusted. That is, the performance of the model is evaluated on the validation set, and the model structure or parameters are adjusted as needed. During the training process, special loss functions and accuracy calculation methods are also used, such as masked_loss and masked_acc, which can ignore the impact of the padding part and improve the efficiency of training. In addition, the vocabulary data volume and the expected loss and accuracy are also considered during model training to evaluate the initial performance of the model.

[0131] In order to obtain analysis results that better meet the needs of investigation and research, the AI ​​analysis model can be enhanced in specific areas based on the survey data requirements. For example, when the survey tends to be in the direction of social sciences, the proportion of survey literature data, professional literature data, policy data, laws and regulations, etc. included in the training data set of the AI ​​analysis model during the training phase will be relatively increased. This will strengthen the data analysis capabilities of the AI ​​analysis model in survey applications.

[0132] like Figure 7As shown, in some embodiments, in order to obtain answer data, the survey data processing system may call the AI ​​analysis model. The AI ​​analysis model may include a natural language processing module and an information extraction module. The question data is then converted into semantic information by the natural language processing module, and the survey project data is determined based on the semantic information. The information extraction module is then used to extract key information associated with the survey project data from the data analysis source, and the answer data is generated based on the extracted key information and the question data.

[0133] It can be seen that the AI ​​analysis model can realize the intelligence and automation of literature research through artificial intelligence technology, large model technology, automation and data analysis technology, and improve the efficiency of users' literature research. And through natural language processing technology, the literature is classified, clustered and associated to help users quickly find literature related to the research topic. At the same time, machine learning technology can also be used to analyze and predict users' search behavior and provide more personalized literature recommendations. Large model technology and knowledge graph technology can also be used to integrate and present the knowledge in the literature to help users better understand and use the information in the literature. In addition, the AI ​​analysis model can also support the unified output of API interfaces with different rules in multiple scenarios to improve practicality.

[0134] After the question data is input into the AI ​​analysis model, the AI ​​analysis model can use the Retrieval-Augmented Generation (RAG) technology to retrieve knowledge information based on the extracted information in the question data. RAG enhancement technology is an artificial intelligence technology that combines a generative language model and a retrieval system. RAG enhancement technology improves the accuracy and reliability of the model by leveraging the understanding and generation capabilities of a large language model (LLM) trained with large-scale data sets, combined with relevant documents or information provided by the retrieval system.

[0135] In order to perform information retrieval, the survey data processing system can construct a retrieval vector library based on the literature data. That is, in some embodiments, the survey data processing system can obtain literature data and perform text cleaning and word segmentation processing on the literature data to generate a word set. Then, based on the word embedding algorithm, the words in the word set are converted into word vectors of fixed dimensions, and the word vector sequence is obtained by obtaining context information and performing word vector enhancement on the word vector based on the context information. Then, a neural network is used to extract semantic information from the word vector sequence to generate the literature data feature vector. Among them, the feature vector includes a word granularity feature vector, a topic granularity feature vector, and a character granularity feature vector.

[0136] After generating the feature vector of the document data, feature fusion and pooling processing can also be performed on the feature vector, wherein feature fusion can combine the feature vectors of word granularity, topic granularity and character granularity, and fuse information of different granularities through fusion mechanisms such as fusion gate mechanism to obtain a richer and more comprehensive text vector representation. Pooling processing uses pooling strategies such as maximum pooling and average pooling to reduce the dimension and abstract the feature vector to obtain a fixed-length representation of the text, that is, to generate a fixed-length text vector representation for subsequent processing. The generated specific number of feature vectors can construct a vector database. Obviously, the vector database is a database associated with the AI ​​analysis model for storing feature vectors associated with document data.

[0137] After constructing the vector database, the AI ​​analysis model can perform information retrieval in the vector database based on the question data to obtain answer data. That is, in some embodiments, after obtaining the question data, the AI ​​analysis model can perform word segmentation processing on the question data to obtain index data. The index data is used to perform information retrieval, so the index data at least includes a first type of keywords used to characterize the survey project data and a second type of keywords used to characterize the analysis intent.

[0138] Then, a query vector is generated based on the index data, and the query vector is compared with the feature vector of the document data stored in the vector database to obtain relevant knowledge information. Corresponding to the question data, the relevant knowledge information may include the document information corresponding to the research project data and the analysis result information corresponding to the analysis intention.

[0139] After acquiring the relevant knowledge information, the answer data can be generated based on the relevant knowledge information. In order to generate the answer data, the AI ​​analysis model can extract key information from the relevant knowledge information and set the information points and logical order of the key information. The key information at least includes literature information corresponding to the research project data and analysis result information corresponding to the analysis intention. Based on the key information after setting the information points and logical order, the answer text is generated according to the answer statement template. The answer statement template includes one or more combinations of preset sentence templates, preset rule templates, and machine learning algorithm templates.

[0140] AI analysis models can convert data, information or knowledge into natural language text based on natural language generation (NLG) technology. To this end, a natural language generation module can be included in the AI ​​analysis model. The NLG module can first understand the input question data and extract key information by parsing the question data. Then perform content planning to determine the information points to be conveyed and their logical order. Then use template-based methods, rule-based methods, or a series of machine learning-based methods to generate answer data text.

[0141] In addition, the NLG module can also adjust the language style according to the target audience or specific scenarios, generate formal or informal answer data text, and adapt to the terminology and expressions of specific fields. The generated answer data text can also undergo post-processing steps such as grammar checking, style consistency checking and fact verification to improve its readability and accuracy.

[0142] In some embodiments, in order to obtain the associated data, the research data processing system may record the source document information of the key information, extract keywords from the key information, and match synonyms according to the keywords, perform a document search according to the keywords and / or the synonyms to obtain the associated document information, and generate the associated data according to the extracted source document information and the associated document information.

[0143] Since different survey methods correspond to different determined data analysis sources, the AI ​​analysis model uses different databases for retrieval and knowledge information extraction in the process of generating answer data and related data based on question data, and the corresponding answer data is also different. For example, when the survey method is a full-library survey, the AI ​​analysis model can provide intelligent consulting, document analysis, innovative ideas and other services based on the full-text database of documents. The large model technology can understand and parse the user's natural language questions, and extract relevant information from the C-net document full-text database and policy and regulations database to provide users with concise and clear answers. It can also quickly retrieve documents and policies by asking questions, obtain key information, data, opinions, etc. in the documents, and support knowledge tracing and document recommendations.

[0144] When the research method is a special research, the special literature vector library built based on user information can be called, and the knowledge information stored in the special literature vector library in the form of vectors can be retrieved and extracted according to the question data. Then, the answer data is generated according to the extracted knowledge information to achieve fast question and answer and accurate query on specific topics, while supporting knowledge tracing and related literature recommendation, thereby improving targeted research capabilities and meeting users' needs for knowledge in the subject area.

[0145] When the research method is single-article study, the AI ​​analysis model can analyze the main content and structure of a single document, and automatically generate a summary of the document content, keywords, related questions, etc. By quickly reading the document, you can get innovative ideas for writing. You can also perform AI study on article fragments, including selecting abbreviations, continuing, expanding, and polishing the fragments, reducing the time for manual reading and processing.

[0146] S105: Display the answer data and the associated data based on the interactive dialogue.

[0147] After obtaining the answer data and related data output by the AI ​​analysis model, the survey data processing system can display the answer data and related data. In order to display the answer data and related data, the survey data processing system can obtain the answer data and related data and add a result display view in the survey interface. The result display view can include two display areas, one for displaying the answer data and the other for displaying the related data.

[0148] For example, after the user enters the question data "Please analyze the research status of project B in field A" based on the text input box in the survey interface, the survey interface can display the question data display view view1 above the text input box. After the AI ​​analysis model processes the question data to obtain the answer data, the answer data display view view2 can be added below the question data display view view1. The display size of the answer data display view can be set according to the specific data volume and data type of the answer data. At the same time, below the answer data display view view2, the associated data display view view3 can be added to display the traceability information of the answer data, including the source document name of the answer data, the web link address, etc.

[0149] It can be seen that based on the survey data processing method provided in the above embodiment, the efficiency and accuracy of the survey and research results can be improved and the cost can be reduced. In addition, the survey and research can be made intelligent and automated, the impact of human factors can be reduced, and the data quality and analysis depth can be improved. It can also provide personalized data support and analysis methods to enhance the pertinence and effectiveness of the survey and research. It can also improve the efficiency and security of data processing and analysis, and reduce the risk of data loss or damage. In addition, it can also enhance the flexibility and adaptability of the survey and research system and method, and improve its ability to cope with complex and changing situations.

[0150] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, some embodiments of this application also provide a survey data processing method based on AI analysis, such as Figure 8 As shown, the method includes:

[0151] S201. Display a survey interface.

[0152] The survey data processing system can respond to the survey instruction input by the user and display the survey interface. The survey interface includes an input control and a mode selection control. The input control is used to allow the user to input text content or upload a data file. The mode selection control includes a first control, a second control and a third control. The survey mode represented by the first control is a full-database survey; the survey mode represented by the second control is a special survey; and the survey mode represented by the third control is a single-article study.

[0153] S202: Acquire an operation event input by the user based on the survey interface.

[0154] During the display of the survey interface, the survey data processing system may continuously monitor the operation events on the survey interface. The operation events may include text input events or file upload events based on input controls. The operation events may also include selection (click) events based on the input of the mode selection control. According to the different operation events monitored, the survey data processing system may obtain survey mode information in different ways based on the operation events.

[0155] S203: If the operation event is based on the input control input, obtain user input information, and extract survey method information from the input information.

[0156] By monitoring the operation events in the research interface, when an input event based on the input control is obtained, the research data processing system can obtain the user input information and extract the key information that can characterize the research method in the user input information. For example, the user input information includes an uploaded literature file and text information with the content "Please analyze the content summary of this article", then according to the text information input by the user and the file upload event, it can be determined that the current research method is single article reading.

[0157] The specific method of extracting the survey method information from the input information may be: firstly, semantic understanding is performed on the input information, that is, the input information is converted into semantic information or structured text data through a natural language processing module. Then, key information related to the survey method is extracted from the semantic understanding result, so as to determine the survey method according to the specific key information content, so as to generate the survey method information.

[0158] S204: If the operation event is input based on the mode selection control, generate the survey mode information based on the mode selection control selected by the operation event.

[0159] That is, the survey data processing system can read the information related to the selection control from the operation event, and determine the selection control selected by the operation event. If the selection control selected by the operation event is the first control, it can be determined that the survey method is a full-library survey, and therefore survey method information can be generated to characterize the survey method as a full-library survey. Similarly, if the selection control selected by the operation event is the second control, it can be determined that the survey method is a special survey. If the selection control selected by the operation event is the third control, it can be determined that the survey method is a single-article study.

[0160] After generating the survey method information, the survey data processing system can determine the data analysis source according to the survey data processing method provided in the above embodiment. Then create an interactive dialogue to obtain the question data input by the user. Input the question data into the AI ​​analysis model, so that the AI ​​analysis model extracts the survey project data from the question data, and obtains key information from the data analysis source based on the survey project data, so as to output and display the answer data and related data. The specific processing method will not be repeated here.

[0161] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, some embodiments of this application also provide a survey data processing method based on AI analysis, such as Fig. 9 As shown, the method includes:

[0162] S301, detecting the current user login status.

[0163] Since the AI ​​analysis model needs to output personalized content for different users in different survey methods, the survey data processing system can respond to the survey instructions input by the user and detect the current user login status. Based on the user login status, the corresponding interface jump and data processing are performed.

[0164] In order to detect the current user login status, the survey data processing system can perform session tracking after obtaining the survey instruction, that is, to implement session tracking through HTTP Cookies or Session. When a user visits the survey interface for the first time, the server will generate a unique session ID and store it in the application through HTTP Cookies or other forms. Each time a user visits the website, the server will check the session ID in the request to verify the user's identity and online status.

[0165] In addition to the above methods for detecting user login status, the survey data processing system can also determine the current user login status based on one or more combinations of methods such as Token, setting Cookie, checking user information in Session, keyword matching of request header information, heartbeat mechanism, etc.

[0166] S302: If the login status is logged in, the step of displaying a survey interface is executed.

[0167] By detecting the login status, if the login status is logged in, the survey interface can be directly displayed, and the survey method information can be obtained in the manner provided in the above embodiment, and the data analysis source can be determined according to the survey method information. Then an interactive dialogue is created to obtain the question data input by the user. The question data is input into the AI ​​analysis model to obtain and display the answer data and related data.

[0168] S303: If the login status is not logged in, display the login interface, verify the user information input by the user based on the login interface, and after the user information login verification is passed, perform the step of displaying the survey interface.

[0169] By detecting the login status, if the login status is not logged in, the user login operation needs to be performed. That is, a login interface can be displayed, and the login interface can include a user information input control for the user to enter user information. After the user completes the information input and clicks the login control, the survey data processing system can verify the user information entered by the user based on the login interface, and after the user information login verification is passed, the survey interface is displayed, and the steps of obtaining survey method information and the like are performed in the manner provided in the above embodiment.

[0170] Further, as a specific implementation of the survey data processing method based on AI analysis in the above embodiment, the present application embodiment provides a survey data processing system based on AI analysis, such as Fig.10 As shown, the system includes:

[0171] A survey method acquisition module is configured to acquire survey method information in response to a survey instruction input by a user, wherein the survey method information is used to represent the survey method specified by the user;

[0172] An analysis source module is configured to determine a data analysis source according to the survey method information, wherein the data analysis source is associated with the survey method corresponding to the survey method information;

[0173] A dialogue module is configured to create an interactive dialogue and obtain question data input by the user based on the interactive dialogue;

[0174] The AI ​​analysis module is configured to input the question data into an AI analysis model to obtain answer data and related data output by the AI ​​analysis model, wherein the AI ​​analysis model is a question-answering model trained using research literature training data; the AI ​​analysis model is used to extract research project data from the question data, and to obtain key information from the data analysis source based on the research project data;

[0175] The display module is configured to display the answer data and the associated data based on the interactive dialogue.

[0176] By applying the technical solution of this embodiment, the system can obtain survey method information and determine the data analysis source based on the survey method information. Then create an interactive dialogue to obtain the question data input by the user. The question data is input into the AI ​​analysis model, so that the AI ​​analysis model extracts the survey project data from the question data, and obtains key information from the data analysis source based on the survey project data to output the answer data and related data. Then display the answer data and related data. The system can process the survey data based on artificial intelligence, automation and data analysis technology, effectively improving the processing efficiency and accuracy of the survey data. And according to different survey methods, key information is obtained from different data analysis sources, which can provide personalized data analysis methods and enhance the pertinence and effectiveness of investigations and research.

[0177] It should be noted that for other corresponding descriptions of the functional units involved in the survey data processing system based on AI analysis provided in the embodiment of the present application, reference can be made to the corresponding descriptions in the survey data processing method based on AI analysis in the above embodiment, which will not be repeated here.

[0178] like Fig.11 As shown, the embodiment of the present application also provides a computer device, which can be specifically a personal computer, a server, a network device, etc. The computer device includes a bus, a processor, a memory and a communication interface, and can also include an input and output interface and a display device. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store location information. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps in each method embodiment are implemented.

[0179] Those skilled in the art will appreciate that the structure of the above-mentioned computer device is only a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components, or combine certain components, or have a different arrangement of components.

[0180] In one embodiment, a computer-readable storage medium is provided. The computer-readable storage medium may be non-volatile or volatile, and stores a computer program thereon. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0181] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0182] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0183] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0184] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0185] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A survey data processing method based on AI analysis, characterized in that: The method comprises: In response to a survey instruction input by a user, obtaining survey method information, wherein the survey method information is used to represent the survey method specified by the user; Determine a data analysis source according to the survey method information, wherein the data analysis source has an association relationship with the survey method corresponding to the survey method information; Creating an interactive dialogue, and acquiring question data input by a user based on the interactive dialogue; The question data is input into an AI analysis model to obtain answer data and associated data output by the AI ​​analysis model, wherein the AI ​​analysis model is a question-answering model trained using research literature training data; the AI ​​analysis model is used to extract research project data from the question data, and to obtain key information in the data analysis source based on the research project data; The answer data and the associated data are displayed based on the interactive dialog.

2. The method according to claim 1, characterized in that Get information on research methods, including: Displaying a research interface, the research interface includes an input control and a mode selection control, the mode selection control includes a first control, a second control and a third control, the research mode represented by the first control is a full-library research; the research mode represented by the second control is a special research; the research mode represented by the third control is a single article study; Obtaining an operation event input by a user based on the survey interface; If the operation event is based on the input control input, obtaining user input information, and extracting survey method information from the input information; If the operation event is based on the mode selection control input, the survey mode information is generated based on the mode selection control selected by the operation event.

3. The method according to claim 2, characterized in that The method further comprises: Detect the current user login status; If the login status is logged in, executing the step of displaying the survey interface; If the login status is not logged in, the login interface is displayed, and the user information input by the user based on the login interface is verified. After the user information login verification is passed, the step of displaying the survey interface is executed.

4. The method according to claim 1, characterized in that: Determine the data analysis source based on the survey method information, including: parsing the survey method specified by the user from the survey method information; If the survey method is a full-database survey, determining that the data analysis source is a full-text literature database; If the research method is a special research, determining that the data analysis source is a special document vector library, the special document vector library is used to store knowledge information in the research documents in the form of vectors; If the research method is single article study, the data analysis source is determined to be an input document, and the input document is a document file input by a user or a document link in a document full-text database specified by a user.

5. The method according to claim 1, characterized in that Inputting the question data into the AI ​​analysis model to obtain answer data and related data output by the AI ​​analysis model, including: Calling the AI ​​analysis model, wherein the AI ​​analysis model includes a natural language processing module and an information extraction module; Converting the question data into semantic information through the natural language processing module; Determining the survey project data according to the semantic information; Extracting key information from the data analysis source by the information extraction module, wherein the key information is information in the data analysis source associated with the survey project data; The answer data is generated according to the extracted key information and the question data.

6. The method according to claim 1, characterized in that The method further comprises: Record the source document information of the key information; Extracting keywords from the key information, and matching synonyms according to the keywords; Performing a literature search based on the keywords and / or the synonyms to obtain related literature information; The associated data is generated according to the extracted source document information and the associated document information.

7. The method according to claim 1, characterized in that If the research method is a special research, the method further includes: Get logged-in user information; Create a new subject document vector library according to the user information; Acquire special document information, which includes search documents, uploaded documents and knowledge base. The search documents are documents obtained by performing document search based on the special index input by the user; the uploaded documents are documents uploaded by the user through the research interface; the knowledge base is an information collection composed of knowledge information; Generate a document vector according to the subject document information, wherein the document vector includes multi-dimensional knowledge information extracted from the document information and traceability information associated with the knowledge information; The document vector is stored in the document vector library.

8. A survey data processing system for AI analysis, characterized in that: The system comprises: A survey method acquisition module is configured to acquire survey method information in response to a survey instruction input by a user, wherein the survey method information is used to represent the survey method specified by the user; An analysis source module is configured to determine a data analysis source according to the survey method information, wherein the data analysis source is associated with the survey method corresponding to the survey method information; A dialogue module is configured to create an interactive dialogue and obtain question data input by the user based on the interactive dialogue; The AI ​​analysis module is configured to input the question data into an AI analysis model to obtain answer data and related data output by the AI ​​analysis model, wherein the AI ​​analysis model is a question-answering model trained using research literature training data; the AI ​​analysis model is used to extract research project data from the question data, and to obtain key information from the data analysis source based on the research project data; The display module is configured to display the answer data and the associated data based on the interactive dialogue.

9. A computer device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images