Information query method and device, computer equipment and storage medium
By constructing and optimizing the language model, the complexity and accuracy of medical insurance knowledge information query are solved, and users' accurate query results in medical insurance insurance are realized, and the efficiency and reliability of information query are improved.
Patent Information
- Application Number
- CN202411206532.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-29
- Publication Date
- 2024-12-13
AI Technical Summary
When querying medical insurance knowledge information, users face problems such as information overload, policy complexity and information accuracy, which leads to increased difficulty in querying and affects the quality of decision-making.
By obtaining medical insurance knowledge information, constructing input information, obtaining related information and linearizing information, inputting it into the language model, calculating the difference between the prediction results and the standard results, optimizing the language model, and obtaining accurate query results.
It realizes accurate and effective query of medical insurance-related knowledge and information, helps users to carry out effective medical insurance insurance operations, and improves the accuracy and reliability of information query.
Smart Images

Figure CN119130674A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image processing technology, specifically to the field of digital medicine, and in particular to an information query method, device, computer equipment and storage medium. Background Art
[0002] With the rapid development of medical insurance business, many users will learn about the information related to medical insurance policies and types in advance before purchasing medical insurance, and then make judgments based on this information, so as to purchase appropriate medical insurance according to the judgment.
[0003] However, when searching for medical insurance knowledge and information, users are facing increasing difficulties and obstacles, mainly due to the rapid development of medical insurance business and the subsequent information explosion. With the continuous improvement and expansion of the medical insurance system, medical insurance-related knowledge and information has shown explosive growth. This information covers many aspects such as medical insurance policies, types, reimbursement ratios, drug catalogs, and the scope of cooperation with medical institutions, making it difficult for users to search and understand.
[0004] The above-mentioned difficulties in querying and understanding are reflected in the following aspects: First, the policies of medical insurance vary from place to place, which requires users to have an in-depth understanding of the specific policies in their area when querying. Due to the diversity and complexity of policies, users often need to filter out the parts related to their actual situation from a lot of information, which undoubtedly increases the difficulty of querying. Secondly, the services launched by various medical insurance companies are different, which provides users with a variety of choices, but also increases the difficulty of comparison and selection. Users need to carefully study the medical insurance plans provided by different companies and compare the differences in their coverage, reimbursement ratio, service quality, etc., in order to make the most suitable decision for themselves. In addition, the difficulty of querying medical insurance knowledge information is also reflected in the accuracy and authority of the information. Due to the complexity of information on the Internet, users often find it difficult to judge which information is true and reliable. Some inaccurate or misleading information may have a negative impact on users' decisions and is not conducive to users' accurate judgment of the acquired medical insurance knowledge information. Summary of the invention
[0005] The purpose of the embodiments of the present application is to propose an information query method, device, computer equipment and storage medium to solve the problem of being unable to accurately and effectively query medical insurance knowledge information to help users perform effective medical insurance purchase operations.
[0006] In order to solve the above technical problems, the present application provides an information query method, which adopts the following technical solution:
[0007] Acquire medical insurance knowledge information, and construct first input information according to the medical insurance knowledge information;
[0008] Acquire associated information according to the first input information to obtain valid medical insurance information;
[0009] Performing information linearization processing on the medical insurance valid information to obtain second input information;
[0010] Inputting the second input information into a preset language model to obtain a predicted medical insurance answer result;
[0011] Calculating the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimizing the language model according to the difference value to obtain an optimized language model;
[0012] Input query information is obtained, and the input query information is input into the optimized language model to obtain a query result.
[0013] Furthermore, the step of obtaining medical insurance knowledge information and constructing the first input information according to the medical insurance knowledge information specifically includes:
[0014] Extract medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information from a preset database;
[0015] Arrange the information relevance between the medical insurance policy document, the medical insurance company information, the medical insurance interpretation information, and the medical insurance content information;
[0016] Constructing a medical insurance knowledge context based on the relevance of the information;
[0017] Extracting context point information of the medical insurance knowledge context;
[0018] The first input information is generated according to the context point information.
[0019] Furthermore, the step of acquiring associated information according to the first input information to obtain valid medical insurance information specifically includes:
[0020] Acquire the first input information;
[0021] Performing word segmentation, part-of-speech tagging, syntactic analysis, and semantic understanding on the first input information to obtain standard first input information;
[0022] Inputting the standard first input information into a pre-trained language learning model to obtain medical insurance policy related information, medical insurance company related information, medical insurance interpretation related information, and medical insurance content related information;
[0023] The information relationships of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information are sorted out and structured to obtain the medical insurance valid information.
[0024] Furthermore, the step of sorting out and structuring the information relationship of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information to obtain the medical insurance valid information specifically includes:
[0025] Obtaining the information relevance and hierarchical relationship of the medical insurance policy related information, the medical insurance company related information, the medical insurance interpretation related information, and the medical insurance content related information;
[0026] Constructing a medical insurance information relationship diagram according to the information relevance and hierarchical relationship;
[0027] Defining fields and relationships of a preset data structure based on the medical insurance information relationship diagram;
[0028] Fill the information fields in the medical insurance information relationship diagram into the preset data structure to obtain medical insurance structure information;
[0029] The medical insurance structure information is subjected to data verification and data cleaning to obtain the medical insurance valid information.
[0030] Furthermore, the step of performing information linearization processing on the medical insurance valid information to obtain the second input information specifically includes:
[0031] Scoring the medical insurance valid information to obtain an information standard score;
[0032] Sorting the information standard scores according to a preset sorting rule to obtain a standard score sorting table;
[0033] Converting the medical insurance valid information into a medical insurance information sequence according to the sorting order of the standard score sorting table;
[0034] The medical insurance information sequence is used as the second input information.
[0035] Furthermore, the step of scoring the effective medical insurance information to obtain an information standard score specifically includes:
[0036] Performing data cleaning and preprocessing on the medical insurance valid information to obtain standard medical insurance valid information;
[0037] Setting weights for information fields corresponding to the standard medical insurance valid information to obtain information field weights;
[0038] The standard medical insurance effective information is weighted and scored according to the preset scoring index and the information field weight to obtain the information standard score.
[0039] Furthermore, the step of calculating the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result specifically includes:
[0040] Dividing the predicted medical insurance answer result and the standard medical insurance answer result into fields to obtain predicted medical insurance key fields and standard medical insurance key fields;
[0041] Perform cross entropy loss calculation based on the predicted medical insurance key field and the standard medical insurance key field to obtain a field loss value;
[0042] An average value of the field loss values is calculated, and the calculated average value is used as the difference value.
[0043] In order to solve the above technical problems, the embodiment of the present application further provides an information query device, which adopts the following technical solution:
[0044] A first information acquisition module, used to acquire medical insurance knowledge information and construct first input information according to the medical insurance knowledge information;
[0045] An associated information acquisition module, used to acquire associated information according to the first input information to obtain valid medical insurance information;
[0046] A second information acquisition module is used to perform information linearization processing on the medical insurance valid information to obtain second input information;
[0047] A result prediction module, used to input the second input information into a preset language model to obtain a predicted medical insurance answer result;
[0048] A model optimization module, used to calculate the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model;
[0049] The result output module is used to obtain input query information, input the input query information into the optimized language model, and obtain the query result.
[0050] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:
[0051] A computer device comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of any one of the above information query methods when executing the computer-readable instructions.
[0052] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:
[0053] A computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the information query method described in any one of the above items are implemented.
[0054] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects: the present embodiment acquires medical insurance knowledge information, and constructs first input information according to the medical insurance knowledge information; acquires associated information according to the first input information to obtain effective medical insurance information; performs information linearization processing on the effective medical insurance information to obtain second input information; inputs the second input information into a preset language model to obtain a predicted medical insurance answer result; calculates the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimizes the language model according to the difference value to obtain an optimized language model; acquires input query information, and inputs the input query information into the optimized language model to obtain a query result. Thus, an optimized language model for accurately querying medical insurance-related knowledge information is effectively constructed, and accurate and effective medical insurance knowledge query results are obtained according to the input query information. The embodiments of the present application can be applied to the medical insurance information search and query interface of the medical insurance business system. Through the embodiments of the present application, medical insurance knowledge information can be accurately and effectively queried to help users perform effective medical insurance operations. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0056] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;
[0057] Figure 2 A flowchart of an embodiment of the information query method according to the present application;
[0058] Figure 3 yes Figure 2 A flowchart of a specific implementation of step S10;
[0059] Figure 4 yes Figure 2 A flowchart of a specific implementation of step S20;
[0060] Figure 5 yes Figure 4 A flowchart of a specific implementation of step S204;
[0061] Figure 6yes Figure 2 A flowchart of a specific implementation of step S30;
[0062] Figure 7 yes Figure 6 A flowchart of a specific implementation of step S301;
[0063] Figure 8 yes Figure 2 A flowchart of a specific implementation of step S50;
[0064] Fig. 9 is a schematic structural diagram of an embodiment of an information query device according to the present application;
[0065] Fig.10 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION
[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.
[0067] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor does it refer to non-related or alternative embodiments that are mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0068] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.
[0069] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.
[0070] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.
[0071] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, desktop computers, etc.
[0072] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .
[0073] It should be noted that the information query method provided in the embodiment of the present application is generally executed by a server, and accordingly, the information query device is generally arranged in the server.
[0074] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.
[0075] Continue to refer Figure 2 , shows a flow chart of an embodiment of a method for information query according to the present application. The information query method comprises the following steps:
[0076] Step S10, obtaining medical insurance knowledge information, and constructing first input information according to the medical insurance knowledge information;
[0077] In this embodiment, the medical insurance knowledge information includes medical insurance policy documents, medical insurance company information, medical insurance interpretation information and medical insurance content information, wherein the above-mentioned medical insurance policy documents, medical insurance company information, medical insurance interpretation information and medical insurance content information are derived from official documents, databases or websites and other information data updated in real time. The first input information is the information field of the medical insurance knowledge information processed by natural language processing technology. The natural language processing technology includes text preprocessing, word embedding, text classification, named entity recognition and other technologies, wherein the text preprocessing includes text cleaning, word segmentation and part-of-speech tagging. Text cleaning is to clean the text information of the medical insurance knowledge information to remove HTML tags, special characters and other messy information that will affect subsequent processing. Word segmentation is to divide the information text corresponding to the medical insurance knowledge information into multiple independent vocabulary units, and part-of-speech tagging is to divide the information text corresponding to the medical insurance knowledge information into multiple independent vocabulary units. It is to determine the part of speech of each divided vocabulary unit in the medical insurance knowledge information; word embedding is the process of mapping words in the medical insurance knowledge information to low-dimensional real number vectors for information vector conversion. The medical insurance knowledge information is processed by word embedding to obtain the medical insurance knowledge information vector; text classification is to classify different categories of texts corresponding to the medical insurance knowledge information, generally through naive Bayes classifiers, support vector machines (SVM), deep learning models, etc. to classify and divide according to the characteristics of the text; named entity recognition can identify the named entities corresponding to the text information based on preset rules matching the text, thereby adding corresponding entity attributes to the text information. After the above processing, the medical insurance knowledge information is converted into word text information with classification characteristics, part of speech tagging, information vector and entity attributes. The word text information is the first input information.
[0078] Step S20, acquiring associated information according to the first input information to obtain valid medical insurance information;
[0079] In this embodiment, the associated information refers to text information that is associated with the first input information. For example, if the first input information is about the reimbursement policy for a certain disease, then the associated information includes the diagnosis criteria, treatment methods, common medicines, etc. of the disease. The medical insurance validity information is the information that is associated and sorted with the first input information and its associated information.
[0080] Step S30, performing information linearization processing on the medical insurance valid information to obtain second input information;
[0081] In this embodiment, information linearization processing refers to converting complex, nonlinear data or information into a simple, linear form. This embodiment obtains simple, linear medical insurance valid information by performing information linearization processing on the medical insurance valid information, and uses the simple, linear medical insurance valid information as the second input information.
[0082] Step S40, inputting the second input information into a preset language model to obtain a predicted medical insurance answer result;
[0083] In this embodiment, the preset language model is a large language model. A large language model (LLMs) is a deep learning model trained using massive data sets. It is based on a deep learning algorithm and uses multiple converter models at the bottom layer. The converter model is usually a neural network composed of an encoder and a decoder with self-attention function. The large language model (LLMs) performs model training based on unsupervised learning technology. It learns the patterns and rules in the second input information and generates new text information through these patterns and rules, thereby effectively outputting valuable predicted medical insurance answer results.
[0084] Step S50, calculating the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimizing the language model according to the difference value to obtain an optimized language model;
[0085] In this embodiment, the standard medical insurance answer result is screened information after judging the validity and value of the medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information obtained above. The information validity judgment and value judgment can be performed by setting preset judgment criteria, or by performing information scoring operations using a machine learning model. The above machine learning model may include a linear regression model, a decision tree regression model, or a random forest regression model.
[0086] Step S60, obtaining input query information, inputting the input query information into the optimized language model, and obtaining a query result;
[0087] In this embodiment, the input query information refers to the actual query information input by the user to the optimized language model, wherein the input query information is obtained in real time, and the query result refers to the result information output by the optimized language model after performing a result search query based on the input query information input by the user.
[0088] In this embodiment, the above method is applied to a medical business system. The medical insurance in the medical business system is one or more of medical insurance, maternity insurance, critical illness insurance, and health insurance. The medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information corresponding to the medical insurance knowledge information are all stored in the medical insurance system and the disease insurance system and obtained by the above systems to perform information query on the knowledge information involved in one or more of the above medical insurances. Specifically, in this embodiment, the information query corresponds to the knowledge information query of medical insurance, and the query result is the query feedback information of the information query interface of the medical business system.
[0089] This embodiment acquires medical insurance knowledge information, constructs first input information based on the medical insurance knowledge information; acquires related information based on the first input information to obtain valid medical insurance information; performs information linearization processing on the valid medical insurance information to obtain second input information; inputs the second input information into a preset language model to obtain a predicted medical insurance answer result; calculates the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimizes the language model based on the difference to obtain an optimized language model; acquires input query information, inputs the input query information into the optimized language model to obtain a query result. In this way, an optimized language model for accurately querying medical insurance-related knowledge information is effectively constructed, and accurate and effective medical insurance knowledge query results are obtained based on the input query information. The embodiment of the present application can be applied to the medical insurance information search and query interface of the medical insurance business system. Through the embodiment of the present application, medical insurance knowledge information can be accurately and effectively queried to help users perform effective medical insurance operations.
[0090] Continue to refer Figure 3 In some optional implementations of this embodiment, step S10 includes the following steps:
[0091] Step S101, extracting medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information from a preset database;
[0092] In this embodiment, the medical insurance policy document refers to the official document on medical insurance issued by the government or relevant institutions, including policy provisions, implementation plans, operating details, etc. These documents provide the legal and policy basis for the operation of medical insurance. Medical insurance policy documents mainly come from official channels such as the official government website, the website of the Medical Insurance Bureau, and the website of the Health Commission; Medical insurance company information refers to information about companies or institutions related to medical insurance business, including company name, business scope, service content, contact information, etc. This information is of guiding significance to the public when choosing a medical insurance company or understanding the medical insurance business. Medical insurance company information can be obtained through officially certified medical insurance company websites, medical insurance business agency websites, government-designated medical insurance service platforms, etc.; Medical insurance interpretation information refers to information that explains and explains medical insurance policy documents, medical insurance company information, etc. Medical insurance interpretation information can come from policy interpretation documents issued by the government or relevant agencies, interpretation articles from professional media, interpretation opinions from experts and scholars, etc. These interpretation information usually combines specific cases and actual conditions to provide the public with more intuitive and easy-to-understand explanations; Medical insurance content information refers to information on the specific content, scope, standards, etc. of medical insurance, such as reimbursement ratio, reimbursement scope, drug catalog, etc. Medical insurance content information mainly comes from medical insurance policy documents issued by the government or relevant agencies, business introductions of medical insurance companies, and public information from medical institutions.
[0093] Step S102, sorting out the information relevance between the medical insurance policy document, the medical insurance company information, the medical insurance interpretation information, and the medical insurance content information;
[0094] In this embodiment, information relevance refers to the dependency and association between medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information. The following is an explanation of the information relevance between medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information. The medical insurance interpretation information depends on the medical insurance policy documents, which is a detailed explanation and description of the policy documents. The medical insurance policy documents directly stipulate the medical insurance content information, such as reimbursement ratio, reimbursement scope, drug catalog, etc. Medical insurance companies need to comply with the provisions of the medical insurance policy documents in the course of their operations, such as business scope, service content, etc. The service content provided by the medical insurance company usually corresponds to the medical insurance content information, such as providing drugs that comply with the medical insurance drug catalog, providing medical services that comply with the reimbursement scope, etc.
[0095] Step S103, constructing a medical insurance knowledge context according to the information relevance.
[0096] In this embodiment, the core layer of the medical insurance knowledge context is the medical insurance policy documents, which include national medical insurance policies, laws, regulations, etc., as well as supplementary policy documents issued by local governments and relevant departments; the second layer of the medical insurance knowledge context is the medical insurance content information, which is the information that the public needs to directly refer to and apply when enjoying medical insurance services. It is directly related to the medical insurance policy documents and is the specific implementation content of the policy; the third layer of the medical insurance knowledge context is the medical insurance company information. The information at this layer is of guiding significance to the public when choosing a medical insurance company or understanding the medical insurance business. Medical insurance companies need to comply with the provisions of the medical insurance policy documents when providing services. Therefore, they are closely related to the medical insurance policy documents and the medical insurance content information; the fourth layer of the medical insurance knowledge context is the information at this layer, which is intended to help the public better understand and apply the medical insurance policy documents and the medical insurance content information. Through specific cases and actual situations, the medical insurance interpretation information provides the public with intuitive and easy-to-understand explanations.
[0097] Step S104, extracting context point information of the medical insurance knowledge context;
[0098] In this embodiment, the context point information refers to the key nodes and elements that constitute the medical insurance knowledge context, for example, the core regulations in the medical insurance policy documents: such as the provisions on basic medical insurance in the "Social Insurance Law of the People's Republic of China", local government policies: supplementary policies and regulations on medical insurance issued by local governments, rules and regulations: various rules and regulations and management methods related to the operation of the medical insurance system; reimbursement ratios in medical insurance content information: including reimbursement ratios of different types of medical insurance such as employee medical insurance and resident medical insurance, reimbursement scope: detailed list of reimbursable items and scopes in outpatient, hospitalization and other medical expenses, drug catalogs: including different categories of medical insurance drug catalogs such as Class A and Class B, service items: covering medical insurance coverage Other medical services and health management items; Company name in medical insurance company information: the name of various companies or institutions that provide medical insurance services; Business scope: the specific medical insurance service content and business scope provided by each medical insurance company; Service quality: including the company's service evaluation, user feedback and other information; Contact information: including address, telephone number, official website and other contact information; Policy interpretation in medical insurance interpretation information: detailed interpretation and analysis of medical insurance policy documents; Case analysis: showing the application and effect of medical insurance policies in actual operations through specific cases; Experts and scholars' views: the views and suggestions of experts and scholars in the medical insurance field on medical insurance policies and systems; FAQs: answers and explanations to common questions of the public in medical insurance, etc. The above context point information can be adjusted according to actual conditions.
[0099] Step S105: Generate the first input information according to the context point information.
[0100] In this embodiment, the format of the first input information is determined according to different specific application scenarios and requirements, and information related to the input requirements is filtered out from the extracted context point information. As the first input information, the above input requirements can be pre-set requirement information, which can be extracted from historical user input data.
[0101] This embodiment extracts medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information from a preset database; sorts out the information relevance between the medical insurance policy documents, the medical insurance company information, the medical insurance interpretation information, and the medical insurance content information; constructs a medical insurance knowledge context according to the information relevance; extracts context point information of the medical insurance knowledge context; and generates the first input information according to the context point information. Thus, the first input information that effectively helps query related information can be generated according to specific information data, so as to facilitate subsequent search for corresponding related information according to the first input information.
[0102] Continue to refer Figure 4 In some optional implementations of this embodiment, step S20 includes the following steps:
[0103] Step S201, obtaining the first input information;
[0104] Step S202, performing word segmentation, part-of-speech tagging, syntactic analysis and semantic understanding on the first input information to obtain standard first input information;
[0105] In this embodiment, the word segmentation processing of the first input information is to cut the first input information into separate words or phrases, the part-of-speech tagging processing of the first input information is to tag each word or phrase with its part of speech, such as noun, verb, adjective, etc., for subsequent processing and analysis, the syntactic analysis of the first input information is to identify the grammatical structure in the sentence and the relationship between the sentence components, which helps to understand the overall meaning of the sentence, and the semantic understanding of the first input information is to further understand the deep meaning of the text on the basis of word segmentation, part-of-speech tagging and syntactic analysis, including the logical relationship and referential relationship between words and sentences. After the first input information undergoes the above processing steps, standard first input information that can be effectively input into the language learning model for training is obtained.
[0106] Step S203, inputting the standard first input information into a pre-trained language learning model to obtain medical insurance policy related information, medical insurance company related information, medical insurance interpretation related information, and medical insurance content related information;
[0107] In this embodiment, the pre-trained language learning model can be a Transformer-based language model, such as BERT, and the above-mentioned language learning model can obtain effective association prediction effects by pre-training a large amount of text data. In this embodiment, the medical insurance policy-related information includes extracting text fragments related to the medical insurance policy, such as policy name, policy content, policy release time, etc., and structuring the information. The medical insurance company-related information includes identifying and organizing information related to the medical insurance company, such as company name, business scope, contact information, etc. The medical insurance interpretation-related information collects and analyzes the interpretation information of the medical insurance policy, medical insurance content, etc., including policy interpretation articles, case analysis, expert opinions, etc. The medical insurance content-related information organizes and summarizes the specific content information of the medical insurance, such as the reimbursement ratio, reimbursement scope, and drug catalog.
[0108] Step S204, sorting out and structuring the information relationship of the medical insurance policy related information, the medical insurance company related information, the medical insurance interpretation related information, and the medical insurance content related information to obtain the medical insurance valid information.
[0109] In this embodiment, labels or keywords corresponding to the information content can be added to the effective medical insurance information to facilitate retrieval and use in subsequent processing.
[0110] This embodiment obtains the first input information; performs word segmentation, part-of-speech tagging, syntactic analysis and semantic understanding on the first input information to obtain standard first input information; inputs the standard first input information into a pre-trained language learning model to obtain medical insurance policy-related information, medical insurance company-related information, medical insurance interpretation-related information, and medical insurance content-related information; sorts out and structures the information relationships of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information, so as to effectively obtain related expansions based on the first input information, and organize the related expanded information into effective medical insurance information with a wider knowledge scope.
[0111] Continue to refer Figure 5 In some optional implementations of this embodiment, step S204 includes the following steps:
[0112] Step S2041, obtaining the information relevance and hierarchical relationship of the medical insurance policy related information, the medical insurance company related information, the medical insurance interpretation related information, and the medical insurance content related information;
[0113] In this embodiment, the above-mentioned information relevance and hierarchical relationship can be extracted from the above-mentioned relationship combing operation and the medical insurance knowledge context.
[0114] Step S2042, constructing a medical insurance information relationship diagram according to the information relevance and hierarchical relationship;
[0115] In this embodiment, a mind map is used to construct a medical insurance information relationship diagram. The medical insurance information relationship diagram should contain nodes (representing information entities) and edges (representing the relationship between information entities), and the relevance and hierarchical relationship of the node information should be clearly identified in the medical insurance information relationship diagram, such as using arrows to indicate the flow and dependency of information.
[0116] Step S2043, defining fields and relationships of a preset data structure based on the medical insurance information relationship diagram;
[0117] In this embodiment, the medical insurance information relationship diagram corresponds to a corresponding data structure, such as a relational database table structure, JSON, XML, etc. By defining each field in the data structure, it is ensured that the information represented by each node and edge in the medical insurance information relationship diagram can be fully stored. By defining the relationship between the fields in the data structure, such as foreign keys, parent-child relationships, etc., the association relationship between the information can be clearly displayed.
[0118] Step S2044, filling the information fields in the medical insurance information relationship diagram into the preset data structure to obtain medical insurance structure information;
[0119] In this embodiment, the medical insurance structure information is obtained by matching each information field in the medical insurance information relationship diagram with the fields in the preset data structure and filling the matching information fields into the preset data structure to ensure the accuracy and completeness of the information.
[0120] Step S2045, performing data verification and data cleaning on the medical insurance structure information to obtain the medical insurance valid information.
[0121] In this embodiment, data verification is to verify whether the data filled in the data structure meets expectations, such as checking whether the field value is legal, whether it meets the constraints, etc., and data cleaning is to clean the data that does not meet the requirements, such as removing duplicate data, correcting erroneous data, etc. In this embodiment, in addition to data verification and data cleaning of the medical insurance structure information, the medical insurance structure information will also be standardized, such as unifying the data format, unit conversion, etc., so as to obtain valid medical insurance information that meets the requirements of subsequent processing.
[0122] This embodiment obtains the information relevance and hierarchical relationship of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information; constructs a medical insurance information relationship graph according to the information relevance and hierarchical relationship; defines the fields and relationships of a preset data structure based on the medical insurance information relationship graph; fills the information fields in the medical insurance information relationship graph into the preset data structure to obtain medical insurance structural information; performs data verification and data cleaning on the medical insurance structural information, so as to effectively obtain medical insurance valid information that is structured and organized for the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information, so as to facilitate subsequent processing.
[0123] Continue to refer Figure 6 In some optional implementations of this embodiment, step S30 includes the following steps:
[0124] Step S301, scoring the medical insurance valid information to obtain an information standard score;
[0125] In this embodiment, the information scoring of the valid medical insurance information is performed according to a preset scoring standard, which at least includes importance: evaluating the value or importance of the information to the user or the system, accuracy: checking the accuracy and reliability of the information, and relevance: measuring the relevance of the information to the user's needs or context. In specific implementation, the valid medical insurance information can also be evaluated for its information importance to obtain the corresponding information standard evaluation, and then sorted according to the information standard evaluation to obtain a standard evaluation sorting table.
[0126] Step S302, sorting the information standard scores according to a preset sorting rule to obtain a standard score sorting table;
[0127] In this embodiment, the preset sorting rule may be to sort from high to low or from low to high according to the standard information score, and a sequence number corresponding to the sorting may be added to the standard score sorting table for sorting display.
[0128] Step S303, converting the medical insurance valid information into a medical insurance information sequence according to the sorting order of the standard score sorting table;
[0129] In this embodiment, the medical insurance information sequence refers to an ordered set processed in a sorting order, and the corresponding valid medical insurance information is sorted in sequence according to the sorting order of the standard score sorting table to obtain the medical insurance information sequence.
[0130] Step S304: using the medical insurance information sequence as the second input information.
[0131] In this embodiment, the valid medical insurance information is scored to obtain the standard information score; the standard information score is sorted according to a preset sorting rule to obtain a standard score sorting table; the valid medical insurance information is converted into a medical insurance information sequence according to the sorting order of the standard score sorting table; and the medical insurance information sequence is used as the second input information. Thus, the second input information that is sorted in sequence according to a certain rule can be effectively obtained.
[0132] Continue to refer Figure 7 In some optional implementations of this embodiment, step S301 includes the following steps:
[0133] Step S3011, performing data cleaning and preprocessing on the medical insurance valid information to obtain standard medical insurance valid information;
[0134] In this embodiment, data cleaning is to clean the data that does not meet the requirements, such as removing duplicate data, correcting erroneous data, etc. The preprocessing includes data verification. By performing data cleaning and data verification on the valid medical insurance information, it is ensured that the obtained standard valid medical insurance information is valuable information.
[0135] Step S3012, setting weights for information fields corresponding to the standard medical insurance validity information to obtain information field weights;
[0136] In this embodiment, the weight can be set according to the importance of each information field. For example, the current information field is compensation, and the weight is set to 0.3; the current information field is insurance, and the weight is set to 0.5. The specific weight setting can be adjusted according to actual needs. The importance of the information field can be obtained by analyzing historical query data.
[0137] Step S3013, weighted scoring is performed on the standard medical insurance validity information according to the preset scoring index and the information field weight to obtain the information standard score.
[0138] In this embodiment, the weighted score can be calculated by summing or averaging the field weights, or by assigning a value to the information field, multiplying the assigned value by the corresponding information field weight, and then summing or averaging the calculated value. The above weighted score calculation method can be adjusted accordingly based on actual conditions.
[0139] This embodiment obtains standard medical insurance valid information by performing data cleaning and preprocessing on the medical insurance valid information; weights are set on information fields corresponding to the standard medical insurance valid information to obtain information field weights; and weighted scoring is performed on the standard medical insurance valid information according to preset scoring indicators and the information field weights, thereby obtaining an information standard score that shows the information value of the medical insurance valid information.
[0140] Continue to refer Figure 8 In some optional implementations of this embodiment, step S50 includes the following steps:
[0141] Step S501, dividing the predicted medical insurance answer result and the standard medical insurance answer result into fields to obtain predicted medical insurance key fields and standard medical insurance key fields;
[0142] In this embodiment, by dividing the predicted medical insurance answer results and the standard medical insurance answer results into fields, key field information is obtained to facilitate subsequent comparison processing.
[0143] Step S502, performing cross entropy loss calculation based on the predicted medical insurance key fields and the standard medical insurance key fields to obtain a field loss value;
[0144] In this embodiment, the cross entropy loss is a loss function commonly used in deep learning and machine learning, which is used to measure the difference between the predicted probability distribution output by the model and the true probability distribution. The cross entropy loss calculation is performed through the corresponding cross entropy loss function, and the cross entropy loss function is expressed as: [L = -[y\log\hat{y}+(1-y)\log(1-\hat{y})]], where (y) is the true label value corresponding to the standard medical insurance key field. For binary classification problems, (y) is usually 0 or 1, and (\hat{y}) is the predicted probability value corresponding to the predicted medical insurance key field. For binary classification problems, the value of (\hat{y}) should be between 0 and 1.
[0145] Step S503: Calculate the average value of the field loss values, and use the calculated average value as the difference value.
[0146] In this embodiment, the predicted medical insurance answer result and the standard medical insurance answer result are divided into fields to obtain predicted medical insurance key fields and standard medical insurance key fields; cross entropy loss is calculated based on the predicted medical insurance key fields and the standard medical insurance key fields to obtain field loss values; the average value of the field loss values is calculated, and the calculated average value is used as the difference value. In this way, the difference value that explains the difference between the predicted result and the standard result can be effectively obtained.
[0147] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).
[0148] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.
[0149] Further references Fig. 9 , as a response to the above Figure 1 The present application provides an embodiment of an information query device, and the device embodiment is similar to Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.
[0150] like Fig. 9 As shown, the information query device 700 described in this embodiment includes: a first information acquisition module 701, a related information acquisition module 702, a second information acquisition module 703, a result prediction module 704, a model optimization module 705, and a result output module 706. Among them:
[0151] The first information acquisition module 701 is used to acquire medical insurance knowledge information and construct first input information according to the medical insurance knowledge information;
[0152] The associated information acquisition module 702 is used to acquire associated information according to the first input information to obtain valid medical insurance information;
[0153] The second information acquisition module 703 is used to perform information linearization processing on the medical insurance effective information to obtain second input information;
[0154] A result prediction module 704, used to input the second input information into a preset language model to obtain a predicted medical insurance answer result;
[0155] The model optimization module 705 is used to calculate the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model;
[0156] The result output module 706 is used to obtain input query information, input the input query information into the optimized language model, and obtain query results.
[0157] This embodiment, by adopting the above-mentioned information query device, can obtain medical insurance knowledge information, construct first input information according to the medical insurance knowledge information; obtain related information according to the first input information to obtain effective medical insurance information; perform information linearization processing on the effective medical insurance information to obtain second input information; input the second input information into a preset language model to obtain a predicted medical insurance answer result; calculate the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model; obtain input query information, input the input query information into the optimized language model to obtain a query result. In this way, an optimized language model for accurately querying medical insurance-related knowledge information is effectively constructed, and accurate and effective medical insurance knowledge query results are obtained according to the input query information. The embodiment of the present application can be applied to the medical insurance information search and query interface of the medical insurance business system. Through the embodiment of the present application, medical insurance knowledge information can be accurately and effectively queried to help users perform effective medical insurance operations.
[0158] To solve the above technical problems, the present application also provides a computer device. Fig.10 , Fig.10 This is a basic structural block diagram of the computer device in this embodiment.
[0159] The computer device 10 includes a memory 81, a processor 82, and a network interface 83 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 8 with components 81-83, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.
[0160] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.
[0161] The memory 81 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 81 can be an internal storage unit of the computer device 8, such as a hard disk or memory of the computer device 8. In other embodiments, the memory 81 can also be an external storage device of the computer device 8, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 8. Of course, the memory 81 can also include both the internal storage unit of the computer device 8 and its external storage device. In this embodiment, the memory 81 is generally used to store the operating system and various application software installed on the computer device 8, such as computer-readable instructions of the information query method, etc. In addition, the memory 81 can also be used to temporarily store various types of data that have been output or are to be output.
[0162] The processor 82 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 82 is generally used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to run computer-readable instructions stored in the memory 81 or process data, such as computer-readable instructions for running the information query method.
[0163] The network interface 83 may include a wireless network interface or a wired network interface. The network interface 83 is generally used to establish a communication connection between the computer device 8 and other electronic devices.
[0164] By adopting the above-mentioned computer device, this embodiment can obtain medical insurance knowledge information, construct first input information according to the medical insurance knowledge information; obtain related information according to the first input information to obtain effective medical insurance information; perform information linearization processing on the effective medical insurance information to obtain second input information; input the second input information into a preset language model to obtain a predicted medical insurance answer result; calculate the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model; obtain input query information, input the input query information into the optimized language model to obtain a query result. In this way, an optimized language model for accurately querying medical insurance-related knowledge information is effectively constructed, and accurate and effective medical insurance knowledge query results are obtained according to the input query information. The embodiment of the present application can be applied to the medical insurance information search and query interface of the medical insurance business system. Through the embodiment of the present application, medical insurance knowledge information can be accurately and effectively queried to help users perform effective medical insurance operations.
[0165] The present application also provides another implementation, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the information query method as described above.
[0166] By adopting the above-mentioned computer-readable storage medium, this embodiment can obtain medical insurance knowledge information, construct first input information according to the medical insurance knowledge information; obtain related information according to the first input information to obtain medical insurance valid information; perform information linearization processing on the medical insurance valid information to obtain second input information; input the second input information into a preset language model to obtain a predicted medical insurance answer result; calculate the difference value between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model; obtain input query information, input the input query information into the optimized language model to obtain a query result. In this way, an optimized language model for accurately querying medical insurance-related knowledge information is effectively constructed, and accurate and effective medical insurance knowledge query results are obtained according to the input query information. The embodiment of the present application can be applied to the medical insurance information search and query interface of the medical insurance business system. Through the embodiment of the present application, medical insurance knowledge information can be accurately and effectively queried to help users perform effective medical insurance operations.
[0167] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.
[0168] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.
Claims
1. An information query method, characterized in that: The steps include: Acquire medical insurance knowledge information, and construct first input information according to the medical insurance knowledge information; Acquire associated information according to the first input information to obtain valid medical insurance information; Performing information linearization processing on the medical insurance valid information to obtain second input information; Inputting the second input information into a preset language model to obtain a predicted medical insurance answer result; Calculating the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimizing the language model according to the difference value to obtain an optimized language model; Input query information is obtained, and the input query information is input into the optimized language model to obtain a query result.
2. The information query method according to claim 1, characterized in that: The step of obtaining medical insurance knowledge information and constructing first input information according to the medical insurance knowledge information specifically includes: Extract medical insurance policy documents, medical insurance company information, medical insurance interpretation information, and medical insurance content information from a preset database; Arrange the information relevance between the medical insurance policy document, the medical insurance company information, the medical insurance interpretation information, and the medical insurance content information; Constructing a medical insurance knowledge context based on the relevance of the information; Extracting context point information of the medical insurance knowledge context; The first input information is generated according to the context point information.
3. The information query method according to claim 1, characterized in that: The step of acquiring associated information according to the first input information to obtain valid medical insurance information specifically includes: Acquire the first input information; Performing word segmentation, part-of-speech tagging, syntactic analysis, and semantic understanding on the first input information to obtain standard first input information; Inputting the standard first input information into a pre-trained language learning model to obtain medical insurance policy related information, medical insurance company related information, medical insurance interpretation related information, and medical insurance content related information; The information relationships of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information are sorted out and structured to obtain the medical insurance valid information.
4. The information query method according to claim 3, characterized in that: The step of sorting out and structuring the information relationship of the medical insurance policy-related information, the medical insurance company-related information, the medical insurance interpretation-related information, and the medical insurance content-related information to obtain the medical insurance valid information specifically includes: Obtaining the information relevance and hierarchical relationship of the medical insurance policy related information, the medical insurance company related information, the medical insurance interpretation related information, and the medical insurance content related information; Constructing a medical insurance information relationship diagram according to the information relevance and hierarchical relationship; Defining fields and relationships of a preset data structure based on the medical insurance information relationship diagram; Fill the information fields in the medical insurance information relationship diagram into the preset data structure to obtain medical insurance structure information; The medical insurance structure information is subjected to data verification and data cleaning to obtain the medical insurance valid information.
5. The information query method according to claim 1, characterized in that: The step of performing information linearization processing on the medical insurance valid information to obtain the second input information specifically includes: Scoring the medical insurance valid information to obtain an information standard score; Sorting the information standard scores according to a preset sorting rule to obtain a standard score sorting table; Converting the medical insurance valid information into a medical insurance information sequence according to the sorting order of the standard score sorting table; The medical insurance information sequence is used as the second input information.
6. The information query method according to claim 5, characterized in that: The step of scoring the effective medical insurance information to obtain an information standard score specifically includes: Performing data cleaning and preprocessing on the medical insurance valid information to obtain standard medical insurance valid information; Setting weights for information fields corresponding to the standard medical insurance valid information to obtain information field weights; The standard medical insurance effective information is weighted and scored according to the preset scoring index and the information field weight to obtain the information standard score.
7. The information query method according to claim 1, characterized in that: The step of calculating the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result specifically includes: Dividing the predicted medical insurance answer result and the standard medical insurance answer result into fields to obtain predicted medical insurance key fields and standard medical insurance key fields; Perform cross entropy loss calculation based on the predicted medical insurance key field and the standard medical insurance key field to obtain a field loss value; An average value of the field loss values is calculated, and the calculated average value is used as the difference value.
8. An information query device, characterized in that: include: A first information acquisition module, used to acquire medical insurance knowledge information and construct first input information according to the medical insurance knowledge information; An associated information acquisition module, used to acquire associated information according to the first input information to obtain valid medical insurance information; A second information acquisition module is used to perform information linearization processing on the medical insurance valid information to obtain second input information; A result prediction module, used to input the second input information into a preset language model to obtain a predicted medical insurance answer result; A model optimization module, used to calculate the difference between the predicted medical insurance answer result and the preset standard medical insurance answer result, and iteratively optimize the language model according to the difference value to obtain an optimized language model; The result output module is used to obtain input query information, input the input query information into the optimized language model, and obtain the query result.
9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the information query method according to any one of claims 1 to 7 when executing the computer-readable instructions.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the information query method according to any one of claims 1 to 7 are implemented.
