Method for monitoring quality of long-term care insurance disability assessments based on large models
By using text transcription and large language model analysis on audio data from the long-term care insurance disability assessment process, the problems of low assessment efficiency and unreliable results have been solved, enabling automated and objective disability level judgment and assessment quality report generation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEOPLE'S INSURANCE COMPANY OF CHINA
- Filing Date
- 2026-01-21
- Publication Date
- 2026-06-05
AI Technical Summary
Current long-term care insurance disability assessments are inefficient and cannot guarantee the objectivity and reliability of the results.
By transcribing audio data from the assessment process and using a large language model to analyze the dialogue between the assessor and the assessed, assessment items, time intervals, and assessment scores are extracted and correlation analysis is performed to generate disability level judgment results and assessment quality reports.
It improves assessment efficiency, reduces subjective bias from human intervention, enhances the traceability and reliability of results, and provides insurance institutions with standardized means of assessment quality monitoring.
Smart Images
Figure CN122155850A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of computer software technology, and specifically to a method for monitoring the quality of long-term care insurance disability assessment based on a large model. Background Technology
[0002] In the insurance field, assessors often need to visit the home of the person being assessed (the applicant), conduct interviews and observations based on an assessment scale, complete the assessment scale according to their disability status, calculate the final score, and obtain the assessment conclusion, i.e., the disability level. The system randomly assigns assessors, and assigns a commercial insurance agent to accompany and supervise the visit; two assessors are randomly assigned to supervise each other; video and audio data are randomly checked manually to evaluate the quality of the assessors' work and determine whether it is qualified; and manual random checks are conducted at the home of the person being assessed to inquire whether the assessors' work quality and attitude meet the standards, thereby evaluating the quality of the assessors' work and determining whether it is qualified.
[0003] This approach is inefficient and cannot guarantee the objectivity of the evaluation results. Summary of the Invention
[0004] This disclosure aims to at least partially address one of the technical problems in the related art.
[0005] Therefore, the purpose of this disclosure is to propose a method, device, electronic device, and storage medium for monitoring the quality of long-term care insurance disability assessment based on a large-scale model. By transcribing audio data during the assessment process into text and analyzing the dialogue between the assessor and the assessed based on a large-scale language model, the system achieves automated monitoring of the long-term care insurance disability assessment process. By extracting assessment items and their corresponding time intervals, assessment scores, and time information, and performing correlation analysis, the system can objectively generate disability level judgment results and assessment quality reports. This method not only improves assessment efficiency and reduces subjective bias from human intervention, but also enhances the traceability and reliability of the results through temporal correlation, providing insurance institutions with a standardized and verifiable means of monitoring assessment quality.
[0006] To achieve the above objectives, the first aspect of this disclosure proposes a method for monitoring the quality of long-term care insurance disability assessment based on a large model, comprising: The audio data to be processed is transcribed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process; Determine the first text corresponding to the evaluator user and the second text corresponding to the evaluated user from the dialogue content text; The first text is analyzed based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item. The second text is analyzed based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each said evaluation item score; Based on the first time interval and the second time interval, a correlation analysis is performed on the first analysis results and the second analysis results to determine the disability level assessment results and the assessment quality report for the insured.
[0007] To achieve the above objectives, the monitoring device for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in the second aspect of this disclosure, includes: The text transcription module is used to transcribe the audio data to be processed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process; The first determining module is used to determine, from the dialogue content text, a first text corresponding to the evaluator user and a second text corresponding to the evaluated user. The first analysis module is used to analyze the first text based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item. The second analysis module is used to analyze the second text based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each evaluation item score; The second determining module is used to perform correlation analysis on the first analysis result and the second analysis result based on the first time interval and the second time interval, so as to determine the disability level judgment result and the assessment quality report of the insured.
[0008] The electronic device proposed in the third aspect of this disclosure includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the monitoring method for monitoring the quality of long-term care insurance disability assessment based on a large model as proposed in the first aspect of this disclosure.
[0009] The fourth aspect of this disclosure provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for monitoring the quality of long-term care insurance disability assessment based on a large model as proposed in the first aspect of this disclosure.
[0010] A fifth aspect of this disclosure provides a computer program product in which, when instructions are executed by a processor, the monitoring method for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in a first aspect of this disclosure, is performed.
[0011] This disclosure provides a method, device, electronic device, and storage medium for monitoring the quality of long-term care insurance disability assessment based on a large-scale language model. The method involves transcribing audio data to be processed into text to obtain dialogue content, wherein the audio data is acquired during the user's disability assessment process. From the dialogue content, a first text corresponding to the assessor user and a second text corresponding to the assessed user are determined. The first text is analyzed based on a large-scale language model to obtain a first analysis result, which indicates at least one assessment item and a first time interval corresponding to each assessment item. The second text is analyzed based on the large-scale language model to obtain a second analysis result, which indicates an assessment score for at least one item and a second time interval corresponding to each assessment item score. A correlation analysis is performed between the first and second analysis results based on the first and second time intervals to determine the insured person's disability level judgment and assessment quality report. Therefore, by transcribing audio data during the assessment process into text and analyzing the dialogue content between the assessor and the assessed based on a large-scale language model, automated monitoring of the long-term care insurance disability assessment process is achieved. By extracting assessment items and their corresponding time intervals, assessment scores, and time information, and performing correlation analysis, the system can objectively generate disability level judgment results and assessment quality reports. This method not only improves assessment efficiency and reduces subjective bias from human intervention, but also enhances the traceability and reliability of results through temporal correlation, providing insurance institutions with a standardized and verifiable means of monitoring assessment quality.
[0012] Additional aspects and advantages of this disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this disclosure. Attached Figure Description
[0013] The above and / or additional aspects and advantages of this disclosure will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, in which: Figure 1 This is a flowchart illustrating a method for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in one embodiment of this disclosure. Figure 2 This is a flowchart illustrating a method for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in another embodiment of this disclosure. Figure 3This is a schematic diagram of the long-term care insurance disability assessment quality monitoring process proposed in this disclosure; Figure 4 This is an interactive diagram of the long-term care insurance disability assessment quality system proposed in this disclosure; Figure 5 This is a schematic diagram of the structure of a monitoring device for long-term care insurance disability assessment quality based on a large model, according to another embodiment of this disclosure. Figure 6 This is a block diagram of an electronic device according to an embodiment of the present application. Detailed Implementation
[0014] Embodiments of this disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are used only to explain this disclosure, and should not be construed as limiting this disclosure. Rather, embodiments of this disclosure include all variations, modifications, and equivalents falling within the spirit and scope of the appended claims.
[0015] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this disclosure are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0016] Figure 1 This is a flowchart illustrating a method for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in one embodiment of this disclosure.
[0017] It should be noted that the execution subject of the monitoring method for the quality of long-term care insurance disability assessment based on a large model in this embodiment is a monitoring device for the quality of long-term care insurance disability assessment based on a large model. This device can be implemented by software and / or hardware. The device can be configured in an electronic device, which may include, but is not limited to, a terminal, a server, etc. For example, the terminal may be a mobile phone, a PDA, etc.
[0018] like Figure 1 As shown, this method for monitoring the quality of long-term care insurance disability assessment based on a large model includes: S101: Transcribe the audio data to be processed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process.
[0019] Among them, long-term care insurance disability assessment is a standardized assessment process for determining the degree of disability (such as the level of self-care ability) of insured individuals who require long-term care services (usually the elderly or disabled). The assessment results are directly related to eligibility and level of insurance compensation.
[0020] Among them, the audio data to be processed is the original voice file recorded during the inquiry and communication between the assessor (the assessor) and the insured person or their family member (the assessed person) at the disability assessment site. It is the core data source that needs to be monitored and analyzed.
[0021] Text transcription is the process of converting audio files containing dialogue into readable, structured text records (i.e., "dialogue content text") using automatic speech recognition technology.
[0022] The dialogue text is the plain text obtained after transcription, which records all the dialogue content between the evaluator and the evaluated party during the evaluation process completely or nearly completely.
[0023] Optionally, in some embodiments, the audio of each user's disability assessment can be uploaded to an audio database after completion; based on preset time intervals, audio data to be processed is scanned from the audio database. Thus, by establishing an audio database and periodically scanning the audio data to be processed, a systematic storage and batch processing of the assessment process is achieved. Automatic audio uploading after each assessment, combined with a data extraction mechanism at preset time intervals, ensures the continuity and timeliness of monitoring. This design facilitates retrospective analysis of historical assessment data and supports efficient processing of large-scale audio, providing a scalable data management foundation for quality control in long-term care insurance.
[0024] The audio database is a dedicated data storage system used to centrally and systematically store all recorded audio data to be processed during the assessment process, along with related metadata (such as assessment ID, time, and insured person information).
[0025] The preset time interval is the time period set by the system for periodically executing monitoring and analysis tasks, such as daily, weekly, or after each batch of assessments. It determines the frequency and timeliness of monitoring.
[0026] In this embodiment of the disclosure, when the audio data to be processed is transcribed into text to obtain the dialogue content text, the audio data can be processed into text, thereby facilitating subsequent analysis based on a large model.
[0027] S102: Determine the first text corresponding to the evaluator user and the second text corresponding to the evaluated user from the dialogue content text.
[0028] Among them, the assessment user refers to the professional (such as assessor, social worker, medical staff) who performs the disability assessment and is mainly responsible for asking questions, guiding, and observing and recording during the dialogue.
[0029] The user being assessed refers to the insured person being assessed, or their family member or caregiver, who is primarily responsible for answering questions about their physical condition and self-care ability during the conversation.
[0030] Among them, the first text and the second text refer to the spoken texts that are separated from the "dialogue content text" and are specifically for the "evaluator user" and the "evaluated user", respectively.
[0031] Optionally, in some embodiments, determining the first text corresponding to the evaluator and the second text corresponding to the evaluated user from the dialogue content text can be achieved by: determining the timbre feature corresponding to each character in the dialogue content text based on the audio data to be processed; classifying and organizing the characters in the dialogue content text according to the timbre features to determine the first text corresponding to the evaluator and the second text corresponding to the evaluated user. Thus, speaker separation is achieved through timbre feature recognition in the dialogue text, accurately distinguishing the text content of the evaluator and the evaluated. This method overcomes the problem of ambiguous speaker identity in traditional transcribed texts, ensuring the relevance of subsequent analysis. Timbre-based classification improves the accuracy of text attribution, providing a reliable data foundation for subsequent role-based analysis, thereby enhancing the accuracy and credibility of the evaluation results analysis.
[0032] Timbre, in particular, refers to the physical attributes of a speech signal that characterize the speaker's individual vocal qualities, such as the shape of the vocal tract and pronunciation habits, forming spectral features. It's like a "fingerprint" of the voice, used to distinguish different speakers.
[0033] In the context of text transcription, "character" generally refers to the basic unit that makes up the "dialogue content text," which can refer to Chinese characters, punctuation marks, or words.
[0034] The classification and organization process involves reorganizing the mixed "dialogue text" based on the speaker's identity identified by "timbre characteristics" and assigning it to the "evaluator user" and the "evaluated user" respectively.
[0035] In this embodiment of the disclosure, when the first text corresponding to the evaluator user and the second text corresponding to the evaluated user are determined from the dialogue content text, the dialogue content text can be split.
[0036] S103: Analyze the first text based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item.
[0037] Among them, the large language model is an advanced artificial intelligence model trained on massive amounts of data, capable of understanding and generating human language. In this method, it acts as an "intelligent analyst," used to understand text semantics, extract key information, and make inferences and judgments based on rules.
[0038] The first analysis result is the structured output obtained after the large language model analyzes the "first text" (evaluator). It mainly includes which "evaluation items" the evaluator involved in the dialogue (such as "eating" and "clothing"), as well as the start and end "time interval" of each item being discussed.
[0039] Among them, assessment items refer to the specific assessment dimensions or activities specified in the disability assessment standard scale, such as basic daily life abilities such as "eating, dressing, washing, toileting, and walking".
[0040] The first time interval, on the audio timeline, is the period between the start and end of the evaluator's discussion of a specific "evaluation item." It is used to pinpoint the location of that topic within the evaluation process.
[0041] In this embodiment of the disclosure, when the first text is analyzed based on a large language model to obtain a first analysis result, the parsing processing of the first text can be realized.
[0042] S104: Analyze the second text based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each evaluation item score.
[0043] The second analysis result is the structured output obtained after the large language model analyzes the "second text" (the evaluated party). It mainly includes the evaluation score of the evaluated party's ability on each "evaluation item" (i.e., "item evaluation score"), as well as the "time interval" in which the dialogue content that provides the basis for these scores is located.
[0044] The project evaluation score is a quantitative score given to the level of ability demonstrated by the evaluated party on a certain "evaluation project" based on predefined scoring standards (such as 0 points representing complete self-reliance and 5 points representing complete dependence).
[0045] The second time interval is the conversation period on the audio timeline during which the evaluated party's answers or performance are related to the score of a certain "assessment item".
[0046] S105: Perform a correlation analysis on the first and second analysis results based on the first and second time intervals to determine the disability level assessment results and the quality report of the insured.
[0047] Association analysis is a data processing method that specifically refers to comparing and matching the "first time interval" and the "second time interval" to determine the temporal correspondence between the question (first text) and the answer / performance (second text) of a certain "assessment item", thereby accurately associating the item with the score.
[0048] The disability level assessment result for the insured is based on the "item assessment score" associated with all "assessment items", and is calculated according to the algorithm stipulated in the insurance terms (such as weighted total score or specific item combination) to arrive at the final disability level conclusion (such as mild, moderate or severe disability).
[0049] The assessment quality report, generated through analysis of the assessment process, evaluates the standardization and reliability of the assessment. Its content may include: whether the assessment items were comprehensive, whether the questions and scoring criteria matched, and whether there were any logical contradictions or time anomalies.
[0050] Optionally, in some embodiments, when performing correlation analysis on the first and second analysis results based on a first and second time interval to determine the disability level assessment result and the assessment quality report for the insured, the process may involve: performing correlation analysis on the first and second analysis results based on the first and second time intervals to determine the assessment item score associated with each assessment item; determining the disability level assessment result for the insured based on the assessment item scores associated with different assessment items; determining the description of each assessment item in the first text; and determining the assessment quality report based on the descriptions associated with different assessment items. By associating time intervals, the system can verify the correspondence between scores and assessment descriptions, thereby identifying potential gaps or inconsistencies in the assessment process. The quality report generated based on the descriptions can reveal the completeness and standardization of the assessment, providing data-driven decision support for improving the assessment process and ensuring the fairness of the results.
[0051] In this embodiment, the audio data to be processed is transcribed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process. From the dialogue content text, a first text corresponding to the assessor user and a second text corresponding to the assessed user are determined. The first text is analyzed based on a large language model to obtain a first analysis result, wherein the first analysis result indicates at least one assessment item and a first time interval corresponding to each assessment item. The second text is analyzed based on the large language model to obtain a second analysis result, wherein the second analysis result indicates at least one assessment score and a second time interval corresponding to each assessment score. A correlation analysis is performed on the first and second analysis results based on the first and second time intervals to determine the insured person's disability level judgment result and assessment quality report. Thus, by transcribing the audio data during the assessment process into text and analyzing the dialogue content between the assessor and the assessed person based on a large language model, automated monitoring of the long-term care insurance disability assessment process is achieved. By extracting assessment items and their corresponding time intervals, assessment scores, and time information, and performing correlation analysis, the system can objectively generate disability level judgment results and assessment quality reports. This method not only improves assessment efficiency and reduces subjective bias from human intervention, but also enhances the traceability and reliability of results through temporal correlation, providing insurance institutions with a standardized and verifiable means of monitoring assessment quality.
[0052] Figure 2 This is a flowchart illustrating a method for monitoring the quality of long-term care insurance disability assessment based on a large model, as proposed in another embodiment of this disclosure.
[0053] like Figure 2 As shown, this method for monitoring the quality of long-term care insurance disability assessment based on a large model includes: S201: Transcribe the audio data to be processed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment.
[0054] S202: Determine from the dialogue content text the first text corresponding to the evaluator user and the second text corresponding to the evaluated user.
[0055] For a detailed description of S201 and S202, please refer to the above embodiments, which will not be repeated here.
[0056] S203: Analyze the first text based on a large language model to obtain reference analysis results.
[0057] Among them, the reference analysis results are unstructured results (such as a text summary) that are output by the large language model after performing preliminary semantic analysis on the "first text".
[0058] S204: Extract data from the reference analysis results based on regular expressions to determine at least one evaluation item and a first time interval corresponding to each evaluation item.
[0059] Regular expressions are tools used to describe or match a series of strings that conform to a certain syntax rule. In this method, they serve as a pattern matching rule to accurately extract structured data (such as project names and timestamps) from the text of the "reference analysis results".
[0060] In other words, in this embodiment of the disclosure, the first text can be analyzed based on a large language model to obtain reference analysis results; data extraction is then performed on the reference analysis results based on regular expressions to determine at least one evaluation item and a first time interval corresponding to each evaluation item. Thus, a combination of a large language model and regular expressions is used to extract evaluation items and their time intervals from the evaluation text. The large language model can understand complex expressions in natural language, while regular expressions are used for the precise extraction of structured data. This hybrid approach balances the flexibility of semantic understanding with the standardization of data extraction, ensuring that evaluation items and their time information are accurately and efficiently identified, providing clear structured input for subsequent correlation analysis.
[0061] S205: Analyze the second text based on a pre-set evaluation scale using a large language model to determine at least one item evaluation score and a second time interval corresponding to each evaluation item score.
[0062] The pre-defined assessment scale is a standardized scoring table or set of rules. It specifies the assessment items, the scoring criteria, grading definitions, and corresponding scores for each item. The large language model will analyze and score the "second text" based on this scale.
[0063] In other words, this embodiment of the disclosure can analyze the second text based on a large language model according to a preset evaluation scale to determine at least one item's evaluation score and a second time interval corresponding to each evaluation item's score. Thus, by analyzing the evaluated text using a large language model based on a preset evaluation scale, automated score determination and time interval labeling are achieved. Using the scale as the analysis benchmark ensures the standardization and consistency of score determination, while the recording of time intervals makes the scoring process chronologically traceable. This method improves scoring efficiency, reduces human scoring errors, and provides crucial data for the time dimension analysis of evaluation quality.
[0064] S206: Perform a correlation analysis on the first analysis results and the second analysis results based on the first time interval and the second time interval to determine the disability level judgment result and assessment quality report of the insured.
[0065] For a detailed description of S206, please refer to the above embodiments, which will not be repeated here.
[0066] In this embodiment, the first text is analyzed using a large language model to obtain reference analysis results. Data extraction from the reference analysis results is then performed using regular expressions to determine at least one evaluation item and a first time interval corresponding to each evaluation item. Thus, a combination of a large language model and regular expressions is used to extract evaluation items and their time intervals from the evaluator's text. The large language model can understand complex expressions in natural language, while regular expressions are used for the precise extraction of structured data. This hybrid method balances the flexibility of semantic understanding with the standardization of data extraction, ensuring that evaluation items and their time information are accurately and efficiently identified, providing clear structured input for subsequent correlation analysis. The second text is then analyzed using a large language model based on a pre-set evaluation scale to determine at least one evaluation score and a second time interval corresponding to each evaluation item score. Therefore, by analyzing the evaluated text using a large language model based on a pre-set evaluation scale, automated score determination and time interval labeling are achieved. Using the scale as an analysis benchmark ensures the standardization and consistency of score determination, while the recording of time intervals makes the scoring process chronologically traceable. This method improves scoring efficiency, reduces human scoring errors, and provides key data for time-dimensional analysis of quality assessment.
[0067] In summary, as described in the above embodiments, Figure 3 As shown, Figure 3 This is a schematic diagram of the long-term care insurance disability assessment quality monitoring process proposed in this disclosure, in which the main steps include: Step 01: The operator uploads the evaluation audio.
[0068] Step 02: The program automatically scans and analyzes unanalyzed audio every 5 minutes.
[0069] Step 03: Convert speech to text and obtain the text of the dialogue content of different people based on the voice timbre in the speech. A: Assessor B: The insured person or family member being assessed.
[0070] Step 04: ① The large language model is used to analyze the text content of the evaluators and to obtain the text analysis conclusions.
[0071] ② Use regular expressions to extract a list of relational data (evaluation items, whether mentioned, time interval, judgment criteria) from the analysis conclusions and save it to the database.
[0072] Step 05: Use the large language model to analyze the text content of the insured person or their family member being evaluated, as well as the context, based on the assessment scale, and obtain the scores and time intervals for the assessment items.
[0073] Step 06: Based on the time interval, correlate the analysis conclusions of A and B (the end time of A is the start time of B) to obtain the evaluation items, whether they are mentioned, the time interval, the judgment basis, and the score.
[0074] Step 07: Summarize the assessment results obtained from the audio analysis to produce a disability level judgment and assessment quality report for the insured person.
[0075] Based on the above methods, this disclosure can construct a long-term care insurance disability assessment quality system, such as... Figure 4 As shown, Figure 4 This is an interactive diagram of the long-term care insurance disability assessment quality system proposed in this disclosure.
[0076] Figure 5 This is a schematic diagram of the structure of a monitoring device for long-term care insurance disability assessment quality based on a large model, as proposed in one embodiment of this disclosure.
[0077] like Figure 5 As shown, the monitoring device 50 for monitoring the quality of long-term care insurance disability assessment based on a large model includes: The text transcription module 501 is used to transcribe the audio data to be processed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process. The first determining module 502 is used to determine, from the dialogue content text, a first text corresponding to the evaluator user and a second text corresponding to the evaluated user. The first analysis module 503 is used to analyze the first text based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item. The second analysis module 504 is used to analyze the second text based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each evaluation item score; The second determining module 505 is used to perform correlation analysis on the first analysis result and the second analysis result based on the first time interval and the second time interval, so as to determine the disability level judgment result and the assessment quality report of the insured.
[0078] It should be noted that the aforementioned explanation of the monitoring method for the quality of long-term care insurance disability assessment based on a large model also applies to the monitoring device for the quality of long-term care insurance disability assessment based on a large model in this embodiment, and will not be repeated here.
[0079] In this embodiment, the audio data to be processed is transcribed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process. From the dialogue content text, a first text corresponding to the assessor user and a second text corresponding to the assessed user are determined. The first text is analyzed based on a large language model to obtain a first analysis result, wherein the first analysis result indicates at least one assessment item and a first time interval corresponding to each assessment item. The second text is analyzed based on the large language model to obtain a second analysis result, wherein the second analysis result indicates at least one assessment score and a second time interval corresponding to each assessment score. A correlation analysis is performed on the first and second analysis results based on the first and second time intervals to determine the insured person's disability level judgment result and assessment quality report. Thus, by transcribing the audio data during the assessment process into text and analyzing the dialogue content between the assessor and the assessed person based on a large language model, automated monitoring of the long-term care insurance disability assessment process is achieved. By extracting assessment items and their corresponding time intervals, assessment scores, and time information, and performing correlation analysis, the system can objectively generate disability level judgment results and assessment quality reports. This method not only improves assessment efficiency and reduces subjective bias from human intervention, but also enhances the traceability and reliability of results through temporal correlation, providing insurance institutions with a standardized and verifiable means of monitoring assessment quality.
[0080] According to embodiments of this application, this application also provides an electronic device and a readable storage medium.
[0081] Figure 6 This is a block diagram of an electronic device according to an embodiment of the present application.
[0082] like Figure 6 As shown, the electronic device includes: The memory 601, the processor 602, and the computer instructions stored in the memory 601 and executable on the processor 602.
[0083] When the processor 602 executes instructions, it implements the monitoring method for the quality of long-term care insurance disability assessment based on a large model provided in the above embodiments.
[0084] Furthermore, electronic devices also include: Communication interface 603 is used for communication between memory 601 and processor 602.
[0085] The memory 601 is used to store computer instructions that can be run on the processor 602.
[0086] The memory 601 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.
[0087] The processor 602 is used to implement the monitoring method for the quality of long-term care insurance disability assessment based on a large model in the above embodiments when executing the program.
[0088] If the memory 601, processor 602, and communication interface 603 are implemented independently, then the communication interface 603, memory 601, and processor 602 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0089] Optionally, in a specific implementation, if the memory 601, processor 602, and communication interface 603 are integrated on a single chip, then the memory 601, processor 602, and communication interface 603 can communicate with each other through an internal interface.
[0090] The processor 602 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.
[0091] This application also proposes a computer program product that, when executed by an instruction processor, implements the monitoring method for long-term care insurance disability assessment quality based on a large model, as described in the embodiments of this application.
[0092] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0093] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0094] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.
[0095] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0096] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0097] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
[0098] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
[0099] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.
Claims
1. A method for monitoring the quality of long-term care insurance disability assessment based on a large model, characterized in that, include: The audio data to be processed is transcribed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process; Determine the first text corresponding to the evaluator user and the second text corresponding to the evaluated user from the dialogue content text; The first text is analyzed based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item. The second text is analyzed based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each said evaluation item score; Based on the first time interval and the second time interval, a correlation analysis is performed on the first analysis results and the second analysis results to determine the disability level assessment results and the assessment quality report for the insured.
2. The method as described in claim 1, characterized in that, The method further includes: After each user's disability assessment is completed, the audio of the assessment process is uploaded to the audio database; Based on a preset time interval, the audio data to be processed is determined by scanning the audio database.
3. The method as described in claim 1, characterized in that, Determining the first text corresponding to the evaluator user and the second text corresponding to the evaluated user from the dialogue content text includes: Based on the audio data to be processed, determine the timbre features corresponding to each character in the dialogue content text; The characters in the dialogue text are categorized and organized according to the timbre characteristics to determine the first text corresponding to the evaluator user and the second text corresponding to the evaluated user.
4. The method as described in claim 1, characterized in that, The analysis of the first text based on a large language model to obtain a first analysis result includes: The first text is analyzed based on the large language model to obtain reference analysis results; Data is extracted from the reference analysis results based on regular expressions to determine at least one of the evaluation items and the first time interval corresponding to each of the evaluation items.
5. The method as described in claim 1, characterized in that, The analysis of the second text based on a large language model to obtain a second analysis result includes: Based on the large language model, the second text is analyzed according to a preset evaluation scale to determine at least one of the item evaluation scores, and the second time interval corresponding to each of the evaluation item scores.
6. The method as described in claim 1, characterized in that, The step of performing a correlation analysis on the first analysis results and the second analysis results based on the first time interval and the second time interval to determine the disability level assessment result and the assessment quality report for the insured person includes: A correlation analysis is performed on the first analysis results and the second analysis results based on the first time interval and the second time interval to determine the score of the evaluation item associated with each evaluation item; The disability level of the insured person is determined based on the scores of the assessment items associated with different assessment items. Determine the description of each of the evaluation items in the first text; The assessment quality report is determined based on the descriptions associated with the different assessment items.
7. A monitoring device for the quality of long-term care insurance disability assessment based on a large model, characterized in that, include: The text transcription module is used to transcribe the audio data to be processed into text to obtain the dialogue content text, wherein the audio data to be processed is obtained during the user's disability assessment process; The first determining module is used to determine, from the dialogue content text, a first text corresponding to the evaluator user and a second text corresponding to the evaluated user. The first analysis module is used to analyze the first text based on a large language model to obtain a first analysis result, wherein the first analysis result is used to indicate at least one evaluation item and a first time interval corresponding to each evaluation item. The second analysis module is used to analyze the second text based on a large language model to obtain a second analysis result, wherein the second analysis result is used to indicate at least one item evaluation score and a second time interval corresponding to each evaluation item score; The second determining module is used to perform correlation analysis on the first analysis result and the second analysis result based on the first time interval and the second time interval, so as to determine the disability level judgment result and the assessment quality report of the insured.
8. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.
9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, in, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-6.
10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1-6.