Artificial intelligence-based data processing method and device, computer device and medium

Abstract

The application belongs to the technical field of artificial intelligence, and relates to a data processing method based on artificial intelligence, comprising the following steps: obtaining user data of a target user; performing segmented processing on insurance document data to obtain segmented insurance documents; processing the insurance documents, user health data and product selection data based on a preset summary generation model to generate corresponding summary content; performing clause highlighting processing on the insurance documents based on the user health data and payment method data through a preset highlighting strategy to obtain corresponding first insurance documents; performing optimization processing on the first insurance documents to obtain corresponding second insurance documents; and displaying the summary content and the second insurance documents. The application also provides a data processing device based on artificial intelligence, a computer device and a storage medium. The application can be applied to the data display business scene in the fields of financial technology and medicine, and improves the accuracy, intelligence and adaptability of insurance document display.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology and can be applied to fields such as fintech and healthcare, particularly to data processing methods, devices, computer equipment, and storage media based on artificial intelligence. Background Technology

[0002] In traditional health insurance application services, insurance documents are displayed in a relatively fixed manner, lacking dynamic adjustment and intelligent personalization. Specifically, in the H5 application links currently provided by health insurance companies, the required reading documents, once determined, cannot be updated in real time based on user information. This lack of intelligence and adaptability in document presentation leads to a poor user experience. This crude document display method fails to accurately meet the needs of different users, making it difficult for them to quickly grasp key information while reading application documents, reducing application efficiency and affecting their understanding and selection of insurance products.

[0003] For example, in the auto insurance sector, traditional methods of presenting insurance policy documents treat all users the same, failing to consider crucial factors such as driving experience and vehicle usage frequency. If a user is a novice driver with frequent vehicle use, the document may not highlight high-risk clauses, leading to an incomplete understanding of the insurance coverage and potential claims disputes due to a lack of attention to relevant clauses. In the medical insurance sector, users of different ages and health conditions have significantly different focuses regarding insurance policy terms, and traditional, fixed document presentation methods cannot meet these diverse needs.

[0004] Therefore, there is an urgent need to provide a method for dynamically adjusting insurance document display based on user information, so as to improve the accuracy of insurance document display, meet users' personalized needs, and enhance the overall efficiency of insurance underwriting services. Summary of the Invention

[0005] The purpose of this application is to propose a data processing method, apparatus, computer equipment, and storage medium based on artificial intelligence to solve the technical problem that the display of insurance documents in the existing traditional health insurance underwriting service model lacks intelligence and adaptability.

[0006] Firstly, an artificial intelligence-based data processing method is provided, including: Acquire user data of the target users; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; The insurance document data is segmented to obtain segmented insurance documents; The insurance document, user health data, and product selection data are processed based on a preset summary generation model to generate corresponding summary content. Based on the user's health data and the payment method data, the insurance document is highlighted using a preset highlighting strategy to obtain the corresponding first insurance document. Based on a preset optimization strategy, the first insurance document is optimized for both voice reading and display to obtain the corresponding second insurance document; The summary content and the second insurance document are then displayed and processed.

[0007] Secondly, an artificial intelligence-based data processing device is provided, comprising: The first acquisition module is used to acquire user data of the target user; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; The first processing module is used to segment the insurance document data to obtain segmented insurance documents; The second processing module is used to process the insurance document, the user health data, and the product selection data based on a preset summary generation model to generate corresponding summary content. The third processing module is used to highlight the terms of the insurance document based on the user's health data and the payment method data using a preset highlighting strategy, so as to obtain the corresponding first insurance document. The optimization module is used to perform voice reading optimization and display optimization on the first insurance document based on a preset optimization strategy to obtain the corresponding second insurance document; The display module is used to display the summary content and the second insurance document.

[0008] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described artificial intelligence-based data processing method.

[0009] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the aforementioned artificial intelligence-based data processing method.

[0010] In the aforementioned solution implemented by the artificial intelligence-based data processing method, apparatus, computer equipment, and storage medium, user data of the target user is first acquired. This user data includes user health data, product selection data, payment method data, and insurance document data. The insurance document data is then segmented to obtain segmented insurance documents. Next, the insurance documents, user health data, and product selection data are processed based on a preset summary generation model to generate corresponding summary content. Subsequently, based on the user health data and payment method data, the insurance documents are highlighted using a preset highlighting strategy to obtain a first insurance document. Further, the first insurance document undergoes speech and display optimization based on a preset optimization strategy to obtain a second insurance document. Finally, the summary content and the second insurance document are displayed. Based on the above automated processing flow, this application intelligently generates summary content corresponding to the insurance document by using a summary generation model and combining user health data and product selection data. Simultaneously, based on a highlighting strategy and combining user health data and payment method data, the insurance document is highlighted with relevant clauses. The processed first insurance document is then interactively optimized to obtain a second insurance document. Finally, the summary content and the second insurance document are displayed together. This intelligently provides users with a dynamically adjusted insurance document display method based on user information, effectively improving the accuracy, intelligence, and adaptability of the insurance document display, and effectively meeting users' personalized needs and enhancing the user experience. Attached Figure Description

[0011] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is an exemplary system architecture diagram to which this application can be applied; Figure 2 This is a flowchart of an embodiment of the artificial intelligence-based data processing method according to this application; Figure 3 This is a schematic diagram of a structure of an embodiment of the artificial intelligence-based data processing apparatus according to this application; Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0014] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0016] like Figure 1 As shown, system architecture 100 may include terminal device 101, network 102, and server 103. Terminal device 101 may be a laptop 1011, tablet 1012, or mobile phone 1013. Network 102 is used as a medium to provide a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0017] Users can use terminal device 101 to interact with server 103 via network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0018] Terminal device 101 can be various electronic devices with a display screen and support web browsing. In addition to laptops 1011, tablets 1012, or mobile phones 1013, terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer, and a desktop computer, etc.

[0019] Server 103 can be a server that provides various services, such as a backend server that provides support for the pages displayed on terminal device 101.

[0020] It should be noted that the artificial intelligence-based data processing method provided in the embodiments of this application is generally executed by a server / terminal device, and correspondingly, the artificial intelligence-based data processing device is generally set in the server / terminal device.

[0021] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0022] Continue to refer to Figure 2 The flowchart illustrates an embodiment of the AI-based data processing method according to this application. The order of steps in the flowchart can be changed, and some steps can be omitted, depending on different needs. The AI-based data processing method provided in this application can be applied to any scenario requiring data display, and thus can be applied to products in these scenarios, such as data display products in the financial insurance and digital healthcare fields. The AI-based data processing method includes the following steps: Step S201: Obtain user data of the target user; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data.

[0023] In this embodiment, the artificial intelligence-based data processing method runs on an electronic device (e.g., Figure 1The server / terminal device shown can acquire user data of the target user through wired or wireless connections. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future wireless connection methods. The executing entity of this application is specifically a data processing system, which may be simply referred to as the system.

[0024] The collection of user health data includes: When a user applies for insurance, the system displays a health declaration form. The form presents each data item in a clear and easy-to-understand manner. For example, age and gender use standard text input boxes; body indicators such as BMI, blood pressure, and blood sugar, in addition to text input boxes, also provide reference range prompts; the medical history section includes checkboxes and text supplement areas, making it convenient for users to select chronic disease types and fill in detailed information such as past surgical history; family medical history is also collected through checkboxes or text input. The collected data is stored in a designated table in the database.

[0025] Product selection data collection includes: On the user's insurance product selection interface, basic insurance and various supplementary insurances are displayed in an intuitive list format, with detailed descriptions of each type of insurance, including its features and coverage range. Users select the desired insurance types by checking boxes, and the system records the selected insurance types and corresponding coverage amounts, storing them in relevant data tables.

[0026] Payment method data collection includes: During the payment process, the system offers two options: installment payment and lump-sum payment. If the user chooses installment payment, a pop-up window allows the user to select the number of installments; if the user chooses lump-sum payment, it is recorded directly. This payment method information is stored in conjunction with other user information.

[0027] Insurance document data collection includes: assigning dedicated personnel to collect essential insurance documents, including initial versions and subsequent updates. Document sources may include internal document management systems, external partners, etc. Collected documents are categorized by format and stored in designated folders, while basic document information such as version number and collection date is recorded.

[0028] In addition, the collected user data can be preprocessed: for user health data, outliers can be detected by setting reasonable ranges and rules. For example, a reasonable range can be set for age, and values ​​exceeding this range can be marked as abnormal; for body indicators, whether they are abnormal can be determined according to medical standards, and outliers can be corrected (if there is reasonable basis) or marked. Product selection data and payment method data are converted into a unified encoding format within the system to facilitate subsequent processing and analysis. Regarding document data, specialized document parsing tools are used to extract plain text content from PDF or HTML documents, and formatting is cleaned, removing irrelevant symbols and formatting codes to make it suitable for natural language processing models.

[0029] By comprehensively collecting data on user health, product selection, payment methods, and insurance documents, and preprocessing it, we provide clean, standardized, and usable data for subsequent operations such as dynamic summary generation and dynamic highlighting of terms, ensuring that subsequent processing can be carried out accurately and efficiently.

[0030] Step S202: The insurance document data is segmented to obtain the segmented insurance document.

[0031] In this embodiment, the specific implementation process of segmenting the insurance document data to obtain the segmented insurance document will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0032] Step S203: Based on a preset summary generation model, process the insurance document, the user health data, and the product selection data to generate corresponding summary content.

[0033] In this embodiment, user health data and product selection data are used as important input features and integrated with the segmented insurance document. User health data may include information such as the user's age, gender, height, weight, and past medical history, while product selection data covers the type of insurance product, coverage amount, and coverage period selected by the user. By combining this information with the content of the insurance document, comprehensive input information can be provided to the summary generation model, enabling the model to understand the user's health status and the characteristics of the selected insurance product.

[0034] Then, the integrated input information is fed into the trained summary generation model. The model first encodes the input information, converting user health data, product selection data, and insurance documents into vector representations. For example, each information item in the user health data (such as age, BMI, etc.) is converted into a vector using a separate encoder; similar methods are used for encoding product selection data and insurance documents. Next, the summary generation model uses an attention mechanism to synthesize and analyze these vectors from different sources. The attention mechanism allows the model to focus on the parts of the input information most relevant to the current task. For example, when processing document paragraphs related to underwriting rules, the model will focus on the content related to "underwriting rules for abnormal BMI" in the document, based on the BMI information in the user's health declaration results.

[0035] Subsequently, based on the comprehensive analysis results of the summary generation model, personalized summary content is generated for each insurance document paragraph. The model extracts key information from the insurance document paragraphs based on the user's health status and the characteristics of the selected insurance product, and generates a summary in concise and accurate language. For example, when the system detects that a user's BMI is 30, the model will focus on paragraphs related to "underwriting rules for abnormal BMI" and "supplementary requirements for medical examination reports," extracting key information such as "If BMI exceeds 30, a medical examination report from the past three months is required, including indicators such as blood sugar and blood lipids," and using this information as the summary content to better reflect the user's actual situation and needs.

[0036] Personalized summary generation is the target step in dynamic summary generation. By combining user health data and product selection data with insurance documents, and using a trained summary generation model for comprehensive analysis and summary generation, it is possible to provide users with summary information that is more tailored to their individual circumstances. This personalized summary helps users understand the core content of insurance documents that is relevant to them more quickly and accurately, improving user experience and the processing efficiency of insurance business.

[0037] Step S204: Based on the user's health data and the payment method data, the insurance document is highlighted using a preset highlighting strategy to obtain the corresponding first insurance document.

[0038] In this embodiment, the specific implementation process of highlighting the terms of the insurance document based on the user's health data and the payment method data using a preset highlighting strategy to obtain the corresponding first insurance document will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0039] Step S205: Based on a preset optimization strategy, the first insurance document is subjected to voice reading optimization and display optimization processing to obtain the corresponding second insurance document.

[0040] In this embodiment, the specific implementation process of performing voice reading optimization and display optimization on the first insurance document based on the preset optimization strategy to obtain the corresponding second insurance document will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0041] Step S206: Display the summary content and the second insurance document.

[0042] In this embodiment, the system simultaneously displays the generated summary and the second insurance document with highlighted clauses to the user. This allows the user to quickly grasp the general content of the insurance document by reading the summary, and to rapidly locate specific clauses closely related to their needs using the highlighted sections, thus achieving a more comprehensive and in-depth understanding of the insurance document. For example, in the interactive document interface, a dynamically generated summary is displayed at the top, and the complete insurance document with highlighted clauses is displayed at the bottom. Users can also jump to the corresponding section of the document by clicking on keywords in the summary.

[0043] This application first acquires user data of the target user; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; then, the insurance document data is segmented to obtain segmented insurance documents; subsequently, based on a preset summary generation model, the insurance documents, user health data, and product selection data are processed to generate corresponding summary content; subsequently, based on the user health data and payment method data, the insurance documents are highlighted using a preset highlighting strategy to obtain a corresponding first insurance document; further, based on a preset optimization strategy, the first insurance document is optimized for voice reading and display to obtain a corresponding second insurance document; finally, the summary content and the second insurance document are displayed. Based on the above automated processing flow, this application intelligently generates summary content corresponding to the insurance document by using a summary generation model and combining user health data and product selection data. Simultaneously, based on a highlighting strategy and combining user health data and payment method data, the insurance document is highlighted with relevant clauses. The processed first insurance document is then interactively optimized to obtain a second insurance document. Finally, the summary content and the second insurance document are displayed together. This intelligently provides users with a dynamically adjusted insurance document display method based on user information, effectively improving the accuracy, intelligence, and adaptability of the insurance document display, and effectively meeting users' personalized needs and enhancing the user experience. In some alternative implementations, step S202 includes the following steps: The insurance document data is subjected to feature recognition and keyword extraction to obtain the corresponding target keywords.

[0044] In this embodiment, the feature recognition includes: First, using part-of-speech tagging technology in natural language processing, each word in the insurance document data is tagged with its part of speech, for example, nouns, verbs, and adjectives are marked separately. This is because chapter titles and paragraph topics often contain specific combinations of parts of speech; for example, chapter titles are often noun phrases. Simultaneously, named entity recognition technology is used to identify entities in the document such as personal names, organization names, and specific terms (e.g., insurance-related professional terms such as "health declaration," "underwriting rules," and "claims process"). These entities play an important indicative role in determining the chapter location.

[0045] The keyword extraction described above includes extracting keywords from insurance document data using the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm. TF-IDF is a statistical method used to evaluate the importance of a word to a document in a document set or corpus. The TF-IDF value of each word is calculated based on the TF-IDF algorithm, and words with higher values ​​are selected as keywords. These keywords help us quickly locate key content areas in the document.

[0046] Semantic relationship analysis is performed on the insurance document data to obtain the corresponding target semantic relationships.

[0047] In this embodiment, the semantic relationship analysis includes converting each word in the insurance document data into a vector representation using a word vector model (such as Word2Vec or GloVe). Word vectors can capture the semantic similarity between words. Based on these word vectors, cosine similarity can be used to calculate the semantic similarity between sentences as the corresponding target semantic relationship. Insurance documents are typically structurally complex, containing multiple chapters and paragraphs on different topics.

[0048] The insurance document data is segmented based on the target keywords and the target semantic relationship to obtain the corresponding specified document data.

[0049] In this embodiment, the starting and ending positions of different chapters can be determined by analyzing the semantic similarity between sentences and the semantic association between sentences and keywords. This allows for the analysis of features such as parts of speech, named entities, keywords, and semantic relationships within the text, enabling the insurance document data to be logically segmented into segmented document data. For example, if a series of sentences revolve around the keyword "health disclosure" and exhibit high semantic similarity while clearly distinguishing themselves from the preceding and following content, these sentences can be classified as the "health disclosure" chapter. The determination of paragraph themes is also based on analyzing the semantic concentration of sentences within a paragraph and their association with the overall document theme.

[0050] Use the specified document data as the document data.

[0051] This application obtains corresponding target keywords by performing feature recognition and keyword extraction on the insurance document data; then, it performs semantic relationship analysis on the insurance document data to obtain corresponding target semantic relationships; subsequently, it segments the insurance document data based on the target keywords and target semantic relationships to obtain corresponding designated document data; and finally, it uses the designated document data as the document data. Based on the above processing flow, this application achieves efficient and accurate segmentation of insurance document data into multiple meaningful units by performing feature recognition, keyword extraction, and semantic relationship analysis on insurance document data, and then segments the insurance document data according to the obtained target keywords and target semantic relationships, and uses the obtained designated document data as the corresponding document data. This facilitates the subsequent summary generation model to generate personalized summaries for each paragraph, which helps to improve the accuracy and relevance of the generated summary content.

[0052] In some optional implementations of this embodiment, step S204 includes the following steps: Call the term mapping rule library corresponding to user risk, and the term mapping relationship library corresponding to payment method.

[0053] In this embodiment, the aforementioned clause mapping rule base, also known as the clause-user risk mapping rule base, is established through the following process: Experts and business personnel in the insurance field are organized to conduct in-depth analysis of user health data (such as medical history and physical examination results), insurance behavior (such as selected insurance types and payment methods), and the content of insurance clauses, leveraging their extensive professional knowledge and practical experience. For example, for "chronic disease-specific protection insurance," experts meticulously study the corresponding claims scope, exclusion clauses, and underwriting requirements, clarifying the relationship between these clauses and users who choose this type of insurance, such as the conditions that users with specific chronic diseases need to meet during underwriting. These mapping relationships are then organized into a rule base, stored in a clear structure for easy subsequent querying and use.

[0054] The aforementioned clause mapping database, also known as the payment method-clause mapping database, is established through the following process: Based on different payment methods, such as installment payments and lump-sum payments, professional personnel analyze the clauses that users need to focus on. For example, for installment payment users, the importance of clauses such as installment payment interest rules and explanations of the consequences of late payments is analyzed; for lump-sum payment users, attention is paid to clauses such as the policy effective date after successful payment and invoice issuance rules. These payment methods are then mapped to their corresponding important clauses, forming the payment method-clause mapping database, clearly identifying the clauses that users need to understand under different payment methods.

[0055] Based on the user's health data and the terms mapping rule base, the insurance document is processed by highlighting the terms to obtain the corresponding first processed document.

[0056] In this embodiment, when a user initiates an insurance application, the system acquires the user's health data and insurance behavior data. Then, it performs a matching query in the terms-user risk mapping rule base based on this data. The system compares the user data with the rule conditions one by one according to the logical relationships in the terms-user risk mapping rule base to determine the terms relevant to the user's situation. For example, after the user selects "Chronic Disease Special Protection Insurance," the system finds the relevant claims scope, exclusions, and underwriting requirements for this type of insurance based on the rule base, and records these highlighted terms. Then, the system highlights the recorded terms in the insurance document to obtain the corresponding first processed document.

[0057] The dynamic highlighting of policy terms is designed to emphasize insurance clauses closely related to the user's own situation, enabling the user to quickly locate the content that is important to them. By establishing a policy-user risk mapping rule base, the system can accurately match the clauses that need to be highlighted based on the user's specific information, increasing the user's attention to key clauses and enhancing the user's understanding and awareness of insurance products.

[0058] Based on the payment method data and the terms mapping database, the first processed document is processed by highlighting the terms to obtain the corresponding second processed document.

[0059] In this embodiment, when a user selects a payment method (such as payment method data corresponding to the target user), the system highlights the relevant terms and conditions in the generated summary and insurance document based on the payment method-terms mapping database. Specifically, key information from relevant terms and conditions can be included in the summary during its generation; and in the document display, these terms and conditions can be highlighted using specific styles (such as color and bold font), allowing users to quickly notice important terms related to their payment method and easily understand their payment-related rights and responsibilities.

[0060] Furthermore, the correlation between payment methods and terms aims to help users better understand insurance terms under different payment methods. By establishing a payment method-terms mapping database, the system can accurately display relevant terms based on the user's selected payment method, enabling users to clearly understand their rights and obligations during the payment process, avoiding disputes due to a lack of understanding of the terms, and improving user awareness and satisfaction with the payment process.

[0061] The second processed document is used as the second insurance document.

[0062] This application utilizes a clause mapping rule library corresponding to user risk and a clause mapping relationship library corresponding to payment method. Then, based on the user's health data and the clause mapping rule library, it performs clause highlighting on the insurance document to obtain a first processed document. Subsequently, based on the payment method data and the clause mapping relationship library, it performs clause highlighting on the first processed document to obtain a second processed document. This second processed document is then used as the second insurance document. Based on this processing flow, this application, by combining user health data with the clause mapping rule library to perform clause highlighting on the insurance document, and vice versa, can achieve dynamic highlighting of the insurance document. This intelligently highlights insurance clauses closely related to the user's own situation, allowing users to quickly locate content important to them, thereby increasing user attention to key clauses, enhancing user understanding and awareness of insurance products, effectively improving the intelligence of insurance display and enhancing the user experience.

[0063] In some alternative implementations, step S205 includes the following steps: The first insurance document is modularized into chapters to obtain the corresponding third insurance document.

[0064] In this embodiment, the specific implementation process of modularizing the first insurance document into chapters to obtain the corresponding third insurance document will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0065] Keyword extraction is performed on the abstract content to obtain the corresponding specified keywords.

[0066] In this embodiment, keywords are extracted from dynamically generated summary content using a TF-IDF (Term Frequency-Inverse Document Frequency) algorithm combined with part-of-speech tagging. First, the TF-IDF value of each word in the summary is calculated. Simultaneously, part-of-speech tagging is applied, retaining only nouns, verbs, and other parts of speech with practical meaning, as these are more likely to serve as keywords summarizing the core content of the paragraph. Words with high TF-IDF values ​​and meeting the part-of-speech tagging requirements are selected as the designated keywords.

[0067] Based on the specified keywords, the third insurance document is linked to obtain the corresponding fourth insurance document.

[0068] In this embodiment, the link building process includes: establishing a mapping table between keywords and modular chapters in the system. For each extracted keyword, the most relevant modular chapter is found through string matching or semantic similarity calculation. String matching can directly compare the string similarity between the keyword and the chapter title or content; semantic similarity calculation can utilize the previously mentioned word vector model to convert the keyword and chapter content into vectors, then calculate the cosine similarity between them, select the chapter with the highest similarity as the corresponding chapter of the keyword, and record it in the mapping table. When a user clicks on a keyword in the summary, the system quickly locates and jumps to the corresponding chapter in the document based on the mapping table.

[0069] Keyword extraction and link creation are crucial steps in enabling rapid navigation in interactive documents. Keywords extracted from the summary can encapsulate the core content of paragraphs, providing users with concise information guidance. By establishing links between keywords and modular chapters, when a user is interested in a particular keyword, they can quickly jump to the relevant chapter in the insurance document, saving time spent searching for information and improving the user experience of reading insurance documents.

[0070] The fourth insurance document is optimized for speech reading based on a preset speech synthesis strategy to obtain the corresponding fifth insurance document.

[0071] In this embodiment, the aforementioned speech-to-speech optimization process includes: converting the document text into speech using speech synthesis technology. Then, a suitable speech synthesis engine is selected, such as iFlytek or Baidu Speech Synthesis. These engines typically provide rich interfaces and parameter settings. Before converting the text to speech, the text is preprocessed, including text cleaning (removing special characters, garbled text, etc.) and punctuation processing (adjusting pauses based on punctuation marks, etc.). Then, the preprocessed text is input into the speech synthesis engine, and parameters such as speech rate and intonation are adjusted. Speech rate can be controlled by setting the number of words pronounced per minute, and intonation can be achieved by adjusting the pitch curve of the voice, making the reading effect natural, clear, and consistent with the user's auditory habits.

[0072] The fifth insurance document is then optimized for display to obtain the corresponding sixth insurance document.

[0073] In this embodiment, the above-mentioned display optimization process includes: designing collapse / expand buttons for each chapter and module in the document's interface design. A tree structure is used to represent the relationship between the document's chapters and modules, with each chapter and module serving as a node in the tree. During interface rendering, corresponding HTML elements (such as div elements) are generated based on the tree structure, and collapse / expand buttons are added to each node. When the user clicks the collapse button, the content of the child nodes under that node is hidden using front-end technologies such as JavaScript; when the user clicks the expand button, the content of the child nodes under that node is displayed. Simultaneously, to improve the user experience, corresponding icons can be added to the buttons (such as "+" for expand and "-" for collapse), and animation effects can be added during collapse or expand operations to make the interface changes smoother.

[0074] Among these features, voice reading and multi-level folding / expanding functions are important measures to improve the user experience of reading insurance documents from different aspects. The voice reading function allows users to understand document content by listening when it's inconvenient to read the text, which is especially suitable for users with poor eyesight or those who need to access information while on the move. The multi-level folding / expanding function allows users to control the content they read; they can choose to view or hide parts of the content according to their needs, avoiding information overload and improving reading convenience and efficiency.

[0075] The sixth insurance document is used as the second insurance document.

[0076] This application obtains a corresponding third insurance document by modularizing the first insurance document into chapters; then, it extracts keywords from the summary content to obtain corresponding specified keywords; subsequently, it constructs links in the third insurance document based on the specified keywords to obtain a corresponding fourth insurance document; next, it optimizes the speech reading of the fourth insurance document based on a preset speech synthesis strategy to obtain a corresponding fifth insurance document; further, it optimizes the display of the fifth insurance document to obtain a corresponding sixth insurance document; finally, it uses the sixth insurance document as the second insurance document. Based on the above processing flow, the interactive document structure provided by this application aims to improve the user experience of reading insurance documents. Modular chapter division makes the document structure clearer, keyword extraction and link establishment enable fast navigation, the use of speech synthesis strategy to optimize the speech reading of insurance documents allows users to easily obtain content by listening, and the display optimization of insurance documents allows users to control the reading content independently, enabling users to understand and use insurance documents more efficiently and conveniently.

[0077] In some optional implementations, the modular chapter division of the first insurance document to obtain the corresponding third insurance document includes the following steps: Entity recognition is performed on the first insurance document to obtain the corresponding key entities.

[0078] In this embodiment, named entity recognition (NAME) technology from natural language processing can be used to perform entity recognition on the first insurance document. Specifically, NAME recognition technology can identify various entities in the text. For example, in the "Health Disclosure" section, entities such as "name," "age," and "gender" that belong to the "basic information disclosure" category can be identified; entities such as "height," "weight," and "blood pressure" that belong to the "health indicator disclosure" category can be identified; and various disease names such as "diabetes" and "hypertension" that belong to the "medical history disclosure" category can be identified. By identifying these key entities, the potential content direction of each sub-section can be preliminarily determined.

[0079] The first insurance document is semantically analyzed based on a pre-defined word vector model to obtain the corresponding semantic hierarchical analysis results.

[0080] In this embodiment, the semantic analysis process for the first insurance document includes: using a word vector model (such as Word2Vec or GloVe) to convert each word in the first insurance document's chapters into a vector representation. Word vectors can capture the semantic similarity between words. Based on these word vectors, a clustering algorithm (such as K-Means clustering) is used to cluster the words. The basic idea of ​​K-Means clustering is to divide the data points into K clusters such that the distance from each data point to the center (mean) of its cluster is minimized. Its objective function is... ,in, Let x be the i-th cluster, and x be the cluster. Data points in It is a cluster The center (mean) of the cluster is used. Through cluster analysis, semantically similar words are grouped together to determine the semantic hierarchy of each sub-section, which serves as the corresponding semantic hierarchy analysis result. For example, words related to basic identity information are grouped into one category, corresponding to the "Basic Information Disclosure" sub-section; words related to various physical indicators are grouped into another category, corresponding to the "Physical Indicator Disclosure" sub-section; and words related to medical history are grouped into another category, corresponding to the "Medical History Disclosure" sub-section.

[0081] Based on the key entities and the semantic hierarchy analysis results, the first insurance document is processed to determine the sub-chapter range, thereby obtaining the corresponding sub-chapter range.

[0082] In this embodiment, based on the obtained key entity and semantic hierarchy analysis results, combined with the paragraph structure and sentence coherence of the text, the scope and content of each sub-section of the aforementioned first insurance document are determined. For example, when a piece of text is identified as mainly revolving around basic information such as "name," "age," and "gender," and is semantically coherent and clearly distinguishable from the preceding and following text, this text can be divided into the "Basic Information Disclosure" sub-section. Furthermore, for some cases with ambiguous boundaries, the boundaries of sub-sections can be further clarified by analyzing the semantic similarity and logical relationships between sentences.

[0083] Based on the sub-chapter range, the first insurance document is divided into chapters to obtain the corresponding seventh insurance document.

[0084] In this embodiment, the first insurance document can be divided into sub-sections based on the obtained sub-section range to obtain a seventh insurance document containing multiple divided sub-sections.

[0085] The seventh insurance document is used as the third insurance document.

[0086] This application identifies key entities in the first insurance document; then performs semantic analysis on the first insurance document based on a preset word vector model to obtain corresponding semantic hierarchy analysis results; subsequently, based on the key entities and the semantic hierarchy analysis results, it determines the sub-chapter range of the first insurance document to obtain corresponding sub-chapter ranges; next, it divides the first insurance document into chapters based on the sub-chapter ranges to obtain a corresponding seventh insurance document; finally, it uses the seventh insurance document as the third insurance document. Based on the above processing flow, the modular chapter division proposed in this application is a further refinement of the already segmented document. Since insurance documents are rich and complex, this application, through fine-grained modular chapter division, can make the structure of insurance documents clearer and more organized. This helps users quickly locate content of interest, improves reading efficiency, facilitates users in finding and understanding various information in insurance documents, and thus improves the user experience.

[0087] In some optional implementations of this embodiment, before step S203, the electronic device may further perform the following steps: Collect document data in the target domain and preprocess the document data to obtain corresponding domain document data.

[0088] In this embodiment, the target domain is specifically the insurance domain. A large amount of insurance-related document data is collected, from a wide range of sources, including various insurance product descriptions, terms and conditions, industry reports, etc. The collected document data is then cleaned to remove noise (such as garbled characters and duplicate content) and standardized, for example, by unifying text formatting and word segmentation. Word segmentation is a fundamental step in natural language processing, dividing continuous text into independent lexical units to facilitate model processing. For example, for Chinese text, dictionary-based or statistical-based word segmentation methods can be used.

[0089] Call the preset initial model.

[0090] In this embodiment, the selection process for the initial model may include: comprehensively considering the characteristics of insurance documents and the requirements for summary generation. Insurance documents are characterized by high professionalism, rich terminology, and rigorous logic. Simultaneously, summary generation requires the model to accurately understand the document content and generate concise and accurate summaries. The BERT model is a pre-trained language model based on the Transformer architecture. Pre-trained through Masked Language Modeling (MLM) and Next Sentence Prediction (NSP) tasks, it can learn rich linguistic knowledge and semantic representations, demonstrating excellent performance in handling various natural language understanding tasks. The T5 model, on the other hand, unifies all natural language processing tasks into text-to-text conversion. Through large-scale multi-task pre-training, it possesses powerful language generation capabilities. Given that insurance document summary generation requires both understanding the document content and generating text, models such as BERT or T5 can be selected.

[0091] The initial model is pre-trained based on the domain document data to obtain a trained first generative model.

[0092] In this embodiment, a selected initial model is pre-trained using cleaned and standardized domain document data. During pre-training, the model learns the linguistic features, vocabulary usage, and common expressions in these documents, gradually acquiring a basic understanding of insurance domain language. Taking the BERT model as an example, in the masked language model task, the model randomly masks a portion of the words in the input text and then predicts the masked words based on the context; in the next sentence prediction task, the model determines whether two sentences are adjacent in the original text. Through continuous iterative training, the model continuously adjusts its parameters, improving its understanding and representation capabilities of insurance domain language.

[0093] The first generative model is fine-tuned based on a pre-built summary dataset to obtain a fine-tuned second generative model.

[0094] In this embodiment, the process of constructing the aforementioned summary dataset includes: collecting specific insurance documents from insurance companies and organizing professionals to annotate these documents with summaries. The professionals are then supervised to write concise and accurate summaries based on the document content, ensuring that the summaries cover the core information of the documents. During the annotation process, certain standards and specifications must be followed, such as summary length limits and information completeness. This results in an annotated summary dataset, which will be used for subsequent model fine-tuning.

[0095] The model fine-tuning process includes: fine-tuning the pre-trained first generation model using a labeled summary dataset. The purpose of fine-tuning is to adjust the model's parameters to better adapt to the company's document style and summary generation requirements. During fine-tuning, insurance documents and corresponding summaries are used as input and output, allowing the model to learn how to generate compliant summaries from the documents. By continuously adjusting hyperparameters such as the learning rate and number of iterations, the model's performance is optimized, improving the accuracy and relevance of the summaries until a summary generation model that meets the performance requirements is obtained.

[0096] The second generation model is used as the summary generation model.

[0097] This application collects document data from a target domain and preprocesses the document data to obtain corresponding domain document data. Then, it calls a pre-set initial model; subsequently, it pre-trains the initial model based on the domain document data to obtain a trained first generation model; subsequently, it fine-tunes the first generation model based on a pre-constructed summary dataset to obtain a fine-tuned second generation model; finally, it uses the second generation model as the summary generation model. Based on the above processing flow, this application can provide powerful language processing and generation capabilities for summary generation by selecting a suitable initial model. The pre-training stage allows the model to learn general language knowledge in the insurance field, enabling it to have basic language comprehension capabilities. Fine-tuning based on a summary annotation dataset for specific documents allows the model to better adapt to the company's business needs and document style, generating a summary generation model that more closely matches actual requirements. This allows subsequent use of the summary generation model to effectively improve the accuracy and effectiveness of personalized summary generation.

[0098] In some optional implementations of this embodiment, after step S206, the electronic device may further perform the following steps: During the process of the target user using the second insurance document, user behavior data of the target user is collected.

[0099] In this embodiment, the process of collecting the aforementioned user behavior data includes: 1) Tracking point setup: Carefully setting monitoring points in various chapters and clauses of the second insurance document. For key locations such as chapter titles and important clause paragraphs, tracking points are implemented by embedding specific monitoring code in the HTML elements of the front-end page. For example, in each chapter title... <h2>Tags and clause paragraphs Add data collection scripts within the `<script>` tag. These scripts will be triggered when the user interacts with the page, sending relevant data to the server.

[0100] 2) Click Hotspot Recording: When a user clicks on a location in the document, the embedded code captures the click event and records the coordinates of the click. The system summarizes and analyzes these coordinates to calculate the click frequency of each area. To more intuitively display click hotspots, the document page can be divided into several small grid areas, and the number of clicks within each grid area can be counted. The area with the most clicks is the click hotspot. For example, the document page can be divided into a 100×100 grid, and by counting the number of clicks n_i (where i represents the grid number) within each grid, the grid area with the larger n_i value is identified as the click hotspot.

[0101] 3) Dwell Time Recording: A timer is used to record the duration a user spends on each chapter or clause. The timer starts when a user enters the display area of ​​a chapter or clause; it stops when the user leaves the area and the dwell time is recorded. The entry and exit times can be determined by factors such as whether the user's mouse moves within the area or whether the page is currently displaying the chapter or clause. For example, if the time a user enters chapter A is recorded as t_start and the time they leave is recorded as t_end, then the dwell time T = t_end. t_start.

[0102] 4) Data Association and Storage: Collected user behavior data is associated and stored with user identity information. User identity information, such as user ID and username, can be obtained when a user logs into the system. A dedicated table is created in the database to store user behavior data, containing fields such as user ID, click location coordinates, click time, and dwell time. User behavior data is linked to user identity information through the user ID for subsequent personalized analysis.

[0103] The user behavior data is analyzed using a pre-set data analysis tool to obtain corresponding behavior analysis results.

[0104] In this embodiment, the process of analyzing the aforementioned user behavior data includes: employing various data analysis methods and tools; specifically, using statistical analysis methods to calculate indicators such as the average dwell time and click frequency for each chapter or clause. When calculating the average dwell time, the dwell time of all users on a particular chapter or clause is summed and then divided by the number of users. Click frequency is calculated to understand the degree of user attention to different areas; click frequency f = n / N, where n is the number of clicks on a particular area, and N is the total number of clicks across all areas. Subsequently, based on the statistically obtained click hotspot distribution and dwell time, it is determined which clauses users are more interested in. If a chapter or clause has a high click frequency and a long average dwell time, it indicates that users have a high level of attention to that content; conversely, if the click frequency is low and the average dwell time is short, it indicates that users have a low level of attention to that content. For example, if the click frequency of the "waiting period clause" is 0.8 (assuming a maximum score of 1) and the average dwell time is 30 seconds, while the click frequency of other clauses is mostly between 0.2 and 0.5 and the average dwell time is between 10 and 20 seconds, it can be determined that users are more interested in the "waiting period clause".

[0105] Obtain the preset model adjustment strategy.

[0106] In this embodiment, the strategy for adjusting the model includes: adjusting the parameters and logic of the dynamic summary generation model based on the behavioral analysis results. Dynamic summary generation models typically have adjustable parameters, such as keyword weights and sentence importance assessment thresholds. If a user is found to have viewed a certain clause multiple times, indicating that the clause is important to the user, the system will increase the weight of keywords related to that clause in the summary generation. For example, the keyword "waiting period" in the original "waiting period clause" had a weight of 0.5; after adjustment, its weight is increased to 0.8. Simultaneously, in terms of document display, the highlighting priority of this clause is increased, making it more prominent in the document and easier for users to find quickly.

[0107] Based on the behavioral analysis results, the summary generation model is adjusted accordingly using the model adjustment strategy.

[0108] In this embodiment, the strategy content of the adjustment strategy can be based on the above model, and dynamic adjustment processing of the summary generation model can be performed according to the obtained behavior analysis results.

[0109] This application collects user behavior data from the target user during the use of the second insurance document; then, it analyzes the user behavior data using a preset data analysis tool to obtain corresponding behavior analysis results; subsequently, it obtains a preset model adjustment strategy; and finally, based on the behavior analysis results, it uses the model adjustment strategy to adjust the summary generation model accordingly. Based on this processing flow, this application, through in-depth analysis of the collected user behavior data, can accurately understand the user's level of attention to different terms and content, and then dynamically adjust the summary generation model according to the behavior analysis results. This makes the summary and document display more in line with user needs, improving the efficiency and satisfaction of users in obtaining information, and achieving personalized optimization of interactive documents.

[0110] In some optional implementations of this embodiment, this application also has an insurance document update processing function, the specific implementation process of which includes: Version comparison algorithm application: When insurance documents are updated, version comparison algorithms (such as the Diff algorithm) are used to compare the documents before and after the update. This algorithm identifies the added, modified, and deleted parts by comparing the content of the two versions of the document line by line or character by character. The algorithm analyzes the text structure of the document, finds the differences, extracts these differences, and determines the specific location and scope of the added or modified clauses.

[0111] Update Display: When a user renews their insurance policy, the system will prioritize displaying any newly added or modified terms. The interface design will prominently feature these updates, such as using special color markers or pop-up notifications. Simultaneously, the system will automatically highlight the new or modified terms and provide an explanatory pop-up window detailing the reasons, content, and impact of the updates. This helps users quickly understand the updated document and ensures they are promptly informed of any changes to the insurance terms.

[0112] Among these features, document update processing ensures that users can stay informed about changes to insurance documents. Updated content is accurately extracted through version comparison algorithms and prioritized for display and highlighted when users renew their policies. An explanatory pop-up window is also provided to ensure users clearly understand the updated terms, protecting their right to know and preventing problems during the application or claims process due to a lack of understanding of the updates.

[0113] In some alternative implementations, the user information obtained is subject to user consent and complies with relevant laws and policies.

[0114] Furthermore, any software tools or components not belonging to our company that appear in the embodiments of this application are merely illustrative examples and do not represent actual use.

[0115] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0116] It should be emphasized that, to further ensure the privacy and security of the above-mentioned summary content, the summary content can also be stored in a node of a blockchain.

[0117] The blockchain referred to in this application is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0118] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results. Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0119] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These 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 methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0120] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0121] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of a data processing device based on artificial intelligence, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0122] like Figure 3 As shown, the artificial intelligence-based data processing device 300 described in this embodiment includes: a first acquisition module 301, a first processing module 302, a second processing module 303, a third processing module 304, an optimization module 305, and a processing module 306. Wherein: The first acquisition module 301 is used to acquire user data of the target user; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; The first processing module 302 is used to segment the insurance document data to obtain the segmented insurance document. The second processing module 303 is used to process the insurance document, the user health data, and the product selection data based on a preset summary generation model to generate corresponding summary content. The third processing module 304 is used to perform clause highlighting on the insurance document based on the user health data and the payment method data, and obtain the corresponding first insurance document by using a preset highlighting strategy. The optimization module 305 is used to perform voice reading optimization and display optimization processing on the first insurance document based on a preset optimization strategy to obtain the corresponding second insurance document; The display module 306 is used to display the summary content and the second insurance document.

[0123] In some optional implementations of this embodiment, the first processing module 302 includes: The first extraction submodule is used to perform feature recognition and keyword extraction on the insurance document data to obtain the corresponding target keywords; The analysis submodule is used to perform semantic relationship analysis on the insurance document data to obtain the corresponding target semantic relationship; The first processing submodule is used to segment the insurance document data based on the target keywords and the target semantic relationship to obtain the corresponding specified document data. The first determining submodule is used to use the specified document data as the document data.

[0124] In some optional implementations of this embodiment, the third processing module 304 includes: The calling submodule is used to call the terms mapping rule library corresponding to user risk and the terms mapping relationship library corresponding to payment method; The second processing submodule is used to perform clause highlighting on the insurance document based on the user health data and the clause mapping rule base to obtain the corresponding first processed document; The third processing submodule is used to perform clause highlighting on the first processing document based on the payment method data and the clause mapping relationship library to obtain the corresponding second processing document; The second determining submodule is used to use the second processed document as the second insurance document.

[0125] In some optional implementations of this embodiment, the optimization module 305 includes: The sub-module is used to perform modular chapter division processing on the first insurance document to obtain the corresponding third insurance document; The second extraction submodule is used to extract keywords from the summary content to obtain the corresponding specified keywords; A submodule is constructed to perform link construction processing on the third insurance document based on the specified keywords to obtain the corresponding fourth insurance document; The first optimization submodule is used to perform speech reading optimization processing on the fourth insurance document based on a preset speech synthesis strategy to obtain the corresponding fifth insurance document; The second optimization submodule is used to perform display optimization processing on the fifth insurance document to obtain the corresponding sixth insurance document; The third determining submodule is used to use the sixth insurance document as the second insurance document.

[0126] In some optional implementations of this embodiment, the division into sub-modules includes: The identification unit is used to perform entity identification on the first insurance document to obtain the corresponding key entities; The analysis unit is used to perform semantic analysis on the first insurance document based on a preset word vector model to obtain the corresponding semantic hierarchical analysis results; The processing unit is used to perform sub-chapter range determination processing on the first insurance document based on the key entities and the semantic hierarchy analysis results, so as to obtain the corresponding sub-chapter range; A partitioning unit is used to perform chapter partitioning processing on the first insurance document based on the sub-chapter range to obtain the corresponding seventh insurance document; A determining unit is used to identify the seventh insurance document as the third insurance document.

[0127] In some optional implementations of this embodiment, the artificial intelligence-based data processing device further includes: The first collection module is used to collect document data in the target domain and preprocess the document data to obtain corresponding domain document data. The calling module is used to invoke the preset initial model; The training module is used to pre-train the initial model based on the domain document data to obtain a trained first generative model; The fine-tuning module is used to fine-tune the first generative model based on a pre-built summary dataset to obtain a fine-tuned second generative model. A determination module is used to use the second generation model as the summary generation model.

[0128] In some optional implementations of this embodiment, the artificial intelligence-based data processing device further includes: The second collection module is used to collect user behavior data of the target user during the process of the target user using the second insurance document; The analysis module is used to perform data analysis on the user behavior data based on preset data analysis tools to obtain corresponding behavior analysis results; The second acquisition module is used to acquire preset model adjustment strategies; The adjustment module is used to adjust the summary generation model according to the behavior analysis results and the model adjustment strategy.

[0129] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed] for details. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0130] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing 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 (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0131] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0132] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), 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, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may also include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for data processing methods based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0133] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, for example, to execute computer-readable instructions of the artificial intelligence-based data processing method.

[0134] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0135] Compared with the prior art, the embodiments of this application have the following beneficial effects: In this embodiment, the application intelligently generates summary content corresponding to the insurance document by using a summary generation model and combining user health data and product selection data. Simultaneously, based on a highlighting strategy and combined with user health data and payment method data, the insurance document is highlighted with relevant clauses. The processed first insurance document is then interactively optimized to obtain a second insurance document. Finally, the summary content and the second insurance document are displayed together. This intelligently provides users with a dynamically adjusted insurance document display method based on user information, effectively improving the accuracy, intelligence, and adaptability of the insurance document display, and effectively meeting users' personalized needs and enhancing their user experience.

[0136] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the artificial intelligence-based data processing method described above.

[0137] Compared with the prior art, the embodiments of this application have the following main advantages: In this embodiment, the application intelligently generates summary content corresponding to the insurance document by using a summary generation model and combining user health data and product selection data. Simultaneously, based on a highlighting strategy and combined with user health data and payment method data, the insurance document is highlighted with relevant clauses. The processed first insurance document is then interactively optimized to obtain a second insurance document. Finally, the summary content and the second insurance document are displayed together. This intelligently provides users with a dynamically adjusted insurance document display method based on user information, effectively improving the accuracy, intelligence, and adaptability of the insurance document display, and effectively meeting users' personalized needs and enhancing their user experience.

[0138] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0139] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

[0140] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals. < / h2>

Claims

1. A data processing method based on artificial intelligence, characterized in that, Includes the following steps: Acquire user data of the target users; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; The insurance document data is segmented to obtain segmented insurance documents; The insurance document, user health data, and product selection data are processed based on a preset summary generation model to generate corresponding summary content. Based on the user's health data and the payment method data, the insurance document is highlighted using a preset highlighting strategy to obtain the corresponding first insurance document. Based on a preset optimization strategy, the first insurance document is optimized for both voice reading and display to obtain the corresponding second insurance document; The summary content and the second insurance document are then displayed and processed.

2. The data processing method based on artificial intelligence according to claim 1, characterized in that, The step of segmenting the insurance document data to obtain segmented insurance documents specifically includes: The insurance document data is subjected to feature recognition and keyword extraction to obtain the corresponding target keywords; Semantic relation analysis is performed on the insurance document data to obtain the corresponding target semantic relations; The insurance document data is segmented based on the target keywords and the target semantic relationship to obtain the corresponding specified document data. Use the specified document data as the document data.

3. The data processing method based on artificial intelligence according to claim 1, characterized in that, The step of highlighting the terms of the insurance document based on the user's health data and the payment method data using a preset highlighting strategy to obtain the corresponding first insurance document specifically includes: Call the term mapping rule library corresponding to user risk, and the term mapping relationship library corresponding to payment method; Based on the user's health data and the terms mapping rule base, the insurance document is processed by highlighting the terms to obtain the corresponding first processed document; Based on the payment method data and the terms mapping database, the first processed document is processed by highlighting the terms to obtain the corresponding second processed document; The second processed document is used as the second insurance document.

4. The data processing method based on artificial intelligence according to claim 1, characterized in that, The step of performing voice reading optimization and display optimization on the first insurance document based on a preset optimization strategy to obtain the corresponding second insurance document specifically includes: The first insurance document is modularized into chapters to obtain the corresponding third insurance document; Keyword extraction is performed on the abstract content to obtain the corresponding specified keywords; Based on the specified keywords, the third insurance document is linked to obtain the corresponding fourth insurance document; The fourth insurance document is optimized for speech reading based on a preset speech synthesis strategy to obtain the corresponding fifth insurance document. The fifth insurance document is then optimized for display to obtain the corresponding sixth insurance document; The sixth insurance document is used as the second insurance document.

5. The data processing method based on artificial intelligence according to claim 4, characterized in that, The step of modularizing the first insurance document into chapters to obtain the corresponding third insurance document specifically includes: Entity recognition is performed on the first insurance document to obtain the corresponding key entities; The first insurance document is semantically analyzed based on a pre-defined word vector model to obtain the corresponding semantic hierarchical analysis results. Based on the key entities and the semantic hierarchy analysis results, the first insurance document is processed to determine the sub-chapter range, and the corresponding sub-chapter range is obtained. Based on the sub-chapter range, the first insurance document is divided into chapters to obtain the corresponding seventh insurance document; The seventh insurance document is used as the third insurance document.

6. The data processing method based on artificial intelligence according to claim 1, characterized in that, Before the step of processing the insurance document, the user health data, and the product selection data based on the preset summary generation model to generate the corresponding summary content, the method further includes: Collect document data in the target domain and preprocess the document data to obtain corresponding domain document data; Call the preset initial model; The initial model is pre-trained based on the domain document data to obtain a trained first generative model. The first generative model is fine-tuned based on a pre-built summary dataset to obtain a fine-tuned second generative model. The second generation model is used as the summary generation model.

7. The data processing method based on artificial intelligence according to claim 1 or 6, characterized in that, After the step of displaying the summary content and the second insurance document, the method further includes: During the process of the target user using the second insurance document, user behavior data of the target user is collected; The user behavior data is analyzed using a preset data analysis tool to obtain corresponding behavior analysis results. Obtain the preset model adjustment strategy; Based on the behavioral analysis results, the summary generation model is adjusted accordingly using the model adjustment strategy.

8. A data processing device based on artificial intelligence, characterized in that, include: The first acquisition module is used to acquire user data of the target user; wherein, the user data includes user health data, product selection data, payment method data, and insurance document data; The first processing module is used to segment the insurance document data to obtain segmented insurance documents; The second processing module is used to process the insurance document, the user health data, and the product selection data based on a preset summary generation model to generate corresponding summary content. The third processing module is used to highlight the terms of the insurance document based on the user's health data and the payment method data using a preset highlighting strategy, so as to obtain the corresponding first insurance document. The optimization module is used to perform voice reading optimization and display optimization on the first insurance document based on a preset optimization strategy to obtain the corresponding second insurance document; The display module is used to display the summary content and the second insurance document.

9. A computer device, characterized in that, It includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the data processing method based on artificial intelligence as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the data processing method based on artificial intelligence as described in any one of claims 1 to 7.

Images