Pet traditional chinese medicine diagnosis and treatment service platform based on multi-condition screening and matching method

Abstract

The application relates to the computer technical field and discloses a pet traditional Chinese medicine diagnosis and treatment service platform and a matching method based on multi-condition screening, aiming to solve the problem of diagnosis and treatment resource mismatch caused by the single screening dimension and the lack of standardized matching mechanism of traditional Chinese medicine characteristic attributes in the prior art. The method comprises the following steps: collecting multi-dimensional attribute data of pet physical sign information, traditional Chinese medicine syndrome characteristics, user preferences, and disease types and severity; establishing a standardized veterinary diagnosis and treatment resource attribute system and generating a matching rule library; performing matching calculation by using a weighted matching algorithm or a similarity model; displaying candidate results in order of matching degree, comprehensive score and distance; collecting feedback data to dynamically adjust the weight coefficient and realize algorithm optimization. The system also comprises diagnosis and treatment scheme recommendation, cost estimation and payment docking, and matching closed-loop verification modules. By incorporating traditional Chinese medicine characteristic attributes such as pet physical quality identification, traditional Chinese medicine syndrome classification, and veterinary professional skill level into the screening conditions, the application realizes precise personalized recommendation for pet traditional Chinese medicine diagnosis and treatment, effectively solves the diagnosis and treatment resource mismatch problem, and can realize dynamic evolution and continuous improvement of the matching system accuracy.

Description

Technical Field

[0001] This invention belongs to the field of computer technology, specifically relating to a pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening. Background Technology

[0002] As people's emotional needs for pets continue to increase, the pet medical market continues to expand. Among them, traditional Chinese medicine for pets, with its gentle and unique therapies such as acupuncture, Chinese medicine, and massage, is gradually showing its advantages in treating difficult diseases such as arthritis, chronic skin diseases, and organ degeneration in dogs and cats, and has become an important development direction in the field of pet medical care.

[0003] However, existing pet healthcare resource matching technologies are insufficient to meet the refined needs of pet TCM (Traditional Chinese Medicine) diagnosis and treatment. Specifically, current mainstream pet healthcare recommendation systems or matching methods primarily rely on basic attributes such as pet breed, age, and symptoms to recommend hospitals or veterinarians. Their selection criteria are relatively singular, failing to fully consider key elements unique to pet TCM diagnosis and treatment, such as constitution identification, syndrome classification, and the four diagnostic methods (tongue appearance, pulse, coat color, etc.). This leads to a significant discrepancy between the recommended results and the pet owner's actual treatment needs. More critically, existing technologies lack standardized collection and matching mechanisms for TCM-specific attributes such as professional qualifications, syndrome specialties, and experience. This makes it difficult for pet owners to quickly and accurately locate a TCM veterinarian who truly matches their pet's condition and professional qualifications when faced with a large pool of varying quality pet TCM resources, resulting in resource mismatch and low treatment efficiency. This technological bottleneck has become a core obstacle hindering the standardization and normalization of the pet TCM industry. Summary of the Invention

[0004] The purpose of this invention is to provide a pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening, which can effectively solve the problems in the background art mentioned above.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening, comprising the following specific steps: S1: Collect multi-dimensional attribute data on pet physical signs, TCM syndrome characteristics, user preferences, and disease type and severity to construct a multi-dimensional attribute set of pet medical needs; S2: Establish a standardized TCM veterinary medicine medical resource attribute system, standardizing and mapping elements such as the doctor's professional direction, professional qualifications, syndrome types of expertise, medical experience, fee standards, and user reviews to generate a matching rule base; S3: Use a weighted matching algorithm or similarity model to match the user's medical needs attributes with various server-side resource nodes in the pet TCM database to generate a candidate matching result list; S4: Sort and display the candidate matching results according to dimensions such as matching degree, comprehensive score, and distance; S5: Collect historical service feedback and user satisfaction evaluation data of the matching results, dynamically adjust the weight coefficients of each screening condition, and continuously optimize the matching algorithm.

[0006] Preferably, the pet's physical characteristics information in step S1 includes breed, age, weight, sex, and past medical history; the TCM syndrome characteristics include the results of constitution identification, tongue appearance, pulse appearance, and coat color, etc.; the user preferences include geographical location, price range, appointment time, and doctor qualification level; and the disease type and severity include the specific disease name and the stage of disease development.

[0007] Preferably, the establishment of the traditional Chinese veterinary medicine diagnosis and treatment resource attribute system in step S2 includes the following mapping rules: mapping the doctor's professional direction to common pet disease types; mapping the doctor's professional qualification level to the certification body level; mapping the doctor's specialty syndrome type to the traditional Chinese medicine syndrome differentiation classification system; mapping the doctor's years of clinical experience to the number of successful cases; mapping the fee standard to the diagnosis and treatment service items; and mapping user evaluations to service satisfaction scores.

[0008] Preferably, the weight coefficient configuration of the weighted matching algorithm in step S3 includes: setting corresponding weight coefficients based on the importance of pet physical signs; setting corresponding weight coefficients based on the importance of TCM syndrome characteristics; setting corresponding weight coefficients based on the importance of user preferences; setting corresponding weight coefficients based on the importance of disease type and severity; setting corresponding weight coefficients based on the importance of doctor qualifications and experience; and the sum of all weight coefficients satisfies a preset normalization condition.

[0009] Preferably, the similarity model in step S3 uses the cosine similarity calculation method, which calculates the cosine value of the angle between the user demand attribute vector and the attribute vector of each Chinese veterinary resource node, and generates a similarity value within a preset range. The closer the value is to the preset upper limit, the higher the matching degree.

[0010] Preferably, the priority rule for sorting and displaying in step S4 is as follows: first, sort by matching degree value from high to low; when the matching degree is the same, sort by comprehensive score from high to low; when the comprehensive score is also the same, sort by distance from near to far.

[0011] Preferably, the triggering conditions for dynamically adjusting the weight coefficients in step S5 include: when the matching satisfaction of a certain type of disease is continuously lower than a preset threshold within a preset time period, the weight coefficient of the relevant attributes of that type of disease is automatically increased; when a doctor's historical positive review rate rises to a new level, the weight coefficient of the doctor's qualification attribute is automatically increased; when user feedback indicates that the importance of a certain preference factor has changed, the corresponding weight coefficient is adjusted according to the feedback data.

[0012] Preferably, the present invention also includes a treatment plan recommendation module, which automatically generates a personalized treatment plan for user reference based on the diagnosis opinions of traditional Chinese veterinary physicians in the matching results, combined with the pet's physical condition identification results and current syndrome characteristics. This module includes acupuncture plans, traditional Chinese medicine prescriptions, massage techniques, and dietary therapy suggestions.

[0013] Preferably, the present invention also includes a cost calculation and payment integration module, which automatically calculates the estimated cost based on the recommended treatment plan. The cost details include diagnosis fee, treatment fee, medication fee and follow-up visit fee, and integrates with a third-party payment platform to enable online payment.

[0014] Preferably, the present invention also includes a matching closed-loop verification module, which pushes a satisfaction evaluation questionnaire after the user completes the diagnosis and treatment service, collects evaluation data on dimensions such as diagnosis and treatment effect, service attitude and cost reasonableness, and stores the evaluation results in the database for subsequent weight coefficient optimization.

[0015] Compared with existing technologies, this invention has the following beneficial effects: For the first time, this invention incorporates traditional Chinese medicine (TCM) attributes such as pet constitution identification, TCM syndrome classification, and TCM veterinary professional skill levels into the screening criteria system, achieving precise and personalized recommendations for pet TCM diagnosis and treatment; through a multi-condition screening mechanism, it helps pet owners quickly locate the most suitable doctor and treatment plan from a large pool of pet TCM diagnosis and treatment resources, effectively solving the problem of mismatched treatment resources caused by information asymmetry; through a standardized attribute tagging system, unified screening logic, and transparent cost calculation function, it alleviates to some extent the problems of chaotic pricing and inconsistent service quality in the industry; by collecting subsequent treatment feedback and user evaluation data from the matching results, it iterative optimization of the screening criteria and algorithm weights is achieved, realizing the dynamic evolution of the matching system and continuous improvement in accuracy. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall technical architecture of a pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening; Figure 2 This is a schematic diagram of the core principle framework of the multi-condition weighted matching algorithm and similarity model in this invention; Figure 3 This is a schematic diagram of the logical process framework for collecting pet vital signs information and TCM syndrome characteristics, as well as mapping standardized attributes of diagnostic and treatment resources in this invention. Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow of the matching result sorting and display and dynamic weight adjustment feedback optimization in this invention. Detailed Implementation

[0017] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below. Example 1

[0018] This embodiment applies to a typical application scenario where a pet owner is looking for a suitable traditional Chinese veterinary veterinarian to treat their senior dog suffering from arthritis. In this scenario, the pet owner wants to find a qualified traditional Chinese veterinarian who specializes in treating bone and joint diseases in senior dogs and is located in the local area, while also wanting a clear understanding of the treatment costs. The following sections describe this embodiment in detail from two aspects: system architecture and workflow.

[0019] In terms of system architecture construction, the platform involved in this embodiment includes four core layers: data acquisition layer, matching engine layer, business application layer, and data storage layer. Each layer interacts and communicates with each other through standardized interfaces.

[0020] The data acquisition layer is deployed at the platform's front-end service port. This layer includes four functional units: a pet vital signs information acquisition module, a traditional Chinese medicine syndrome characteristic acquisition module, a user preference setting module, and a disease information entry module. The pet vital signs information acquisition module guides users to input basic pet attribute data through an interactive form interface. This includes the dog's breed information, age calculated from its birth date, weight measurement, sex characteristics, and past medical history. Breed information is categorized into large, medium, and small breeds, age is accurate to the month, and weight is in kilograms. The traditional Chinese medicine syndrome characteristic acquisition module provides an entry point for a constitution identification questionnaire. The system pre-sets nine basic constitution types for users to select and confirm. Tongue information is collected by uploading image files, pulse information is selected by the user based on a given diagram, and coat color information is selected from a drop-down menu to match the actual color. The user preference settings module provides a geographic range selection control, which defaults to setting the search radius centered on the user's current location. The price range setting control supports inputting minimum and maximum prices. The appointment time selection control allows selection of weekdays, weekends, or specific dates. The doctor qualification level checkbox allows selection of levels such as licensed physician, attending physician, and specialist. The disease information entry module provides a drop-down list of common symptoms and a custom input text box, with disease stage options distinguishing between early, middle, and late stages. The raw data collected by each module is formatted and standardized into data packets, then transmitted via an encrypted channel to the matching engine layer for further processing.

[0021] The business application layer is deployed on the platform's backend service port. This layer includes four business functional units: a matching result sorting module, a treatment plan recommendation module, a cost calculation module, and a payment integration module. The matching result sorting module receives the candidate matching list from the matching engine layer and performs secondary sorting according to preset priority rules. First, it checks the matching degree of each candidate result, then sorts them in descending order. For candidates with the same matching degree, it further compares their comprehensive scores and sorts them in descending order of comprehensive score. Finally, for candidates with the same comprehensive score, it calculates the geographical distance and sorts them in ascending order of distance. After sorting, the result data is encapsulated into JSON format data required for frontend rendering. The treatment plan recommendation module is triggered upon successful matching. This module reads the basic diagnostic opinions of the traditional Chinese veterinary physician from the matching results, combines the pet's constitution identification results and current syndrome characteristics data, and uses preset treatment plan generation rules to extract the corresponding acupuncture plan code, traditional Chinese medicine prescription, massage techniques, and dietary recommendations, encapsulating them into a structured treatment plan data package. The cost calculation module uses the service item codes in the treatment plan data package to query the locally maintained service item price list and calculate the total cost. The cost details include four independent cost categories: diagnosis fee, treatment fee, medication fee, and follow-up visit fee. The calculation results are returned to the front end for display to the user. The payment integration module integrates the SDK development package of a third-party payment platform. After the user confirms the cost, the payment interface is invoked to complete the online payment function for the treatment fee.

[0022] The data storage layer adopts a distributed database cluster architecture, comprising three data domains: a pet information database, a traditional Chinese veterinary resource database, and a matching rule database. The pet information database uses a NoSQL document database storage model. Each pet entity's document record includes a unique identifier field, a physical condition information document object, a syndrome characteristic array, and a disease history record array. The system creates an index for each pet entity to support efficient query operations. The traditional Chinese veterinary resource database uses a relational database storage model. The basic information table for traditional Chinese veterinary physicians is linked to the physician qualification certificate table, the specialty syndrome type table, and the clinical experience table via foreign keys. The user evaluation table is linked to the basic information table through the physician identifier field, and the fee standard table is indexed by a composite index of service item number and physician identifier. The matching rule database uses a key-value pair cache database storage model. The weight coefficients, similarity calculation parameters, and dynamic adjustment thresholds for various filtering conditions are mapped and stored in key-value format, supporting instant application of runtime configurations.

[0023] In terms of workflow deduction, the execution process of this embodiment follows the following time sequence logic, and the complete process from the user initiating a medical treatment request to the final completion of the matching result display and payment includes seven core steps.

[0024] The first step is the requirement input and data submission step. Each functional unit of the data acquisition layer presents an interactive interface to guide the user to complete the information filling. After the user completes the input of the pet's vital signs information and clicks the confirmation button, the system verifies the validity of the values ​​in the required fields. After the verification is successful, the data is written to the temporary session storage area. Then, the system continues to guide the user to complete the filling of the next three parts of information. After all the information is filled in, the user clicks the final submission button. The system merges the data from the four modules to generate a classified data message, which is then sent to the matching engine layer through a secure transmission protocol.

[0025] The second step is the requirement parsing and vectorization transformation. After receiving the data message, the preprocessing module of the matching engine layer performs parsing operations, parsing the pet's physical characteristics information into a physical characteristic vector with five component dimensions, the TCM syndrome characteristic information into a syndrome vector with four component dimensions, the user preference information into a preference vector with four component dimensions, and the disease information into a disease vector with two component dimensions. The vector values ​​are standardized according to the pre-set normalization rules. After processing, the four vectors are merged into a single multi-dimensional requirement attribute vector.

[0026] The third step is the resource matching calculation step. The preprocessing module transmits the demand attribute vector to the weighted matching calculation unit. This calculation unit calls the traditional Chinese veterinary medicine physician resource database in the data storage layer to load the attribute vector dataset of all registered traditional Chinese veterinary physicians. For each candidate physician's attribute vector and demand attribute vector, a weighted similarity calculation is performed. The calculation formula adopts the cosine similarity algorithm model, which multiplies and sums the corresponding components of the two vectors and then divides the result by the product of the magnitudes of the two vectors to obtain a similarity value in the range of 0 to 1. The closer the value is to 1, the higher the degree of matching between the physician and the user's needs. The weighting coefficients are set as follows: pet physical characteristics information 0.15, traditional Chinese medicine syndrome characteristics 0.25, user preferences 0.20, disease type and severity 0.25, and physician qualifications and experience 0.15. The sum of the weighting coefficients is strictly equal to 1.0. After the calculation is completed, a candidate matching result list is generated in descending order of similarity value.

[0027] The fourth step is the comprehensive score assignment step. The weighted matching calculation unit transmits the candidate matching result list to the comprehensive score calculation unit. This unit reads the corresponding historical service data record for each candidate result and calculates the comprehensive score assignment based on the doctor's positive review rate, average satisfaction score, and cumulative service times in the user evaluation table. The assignment rule is that the comprehensive score equals the positive review rate multiplied by 60%, plus the average satisfaction score divided by 10 multiplied by 30%, plus the logarithm of service times multiplied by 10%. The calculation result is rounded to two decimal places. After the comprehensive score assignment is completed, the result is sent back to the matching result sorting module.

[0028] The fifth step is the sorting and display step. The matching result sorting module performs a secondary sorting operation, arranging all candidate results in descending order according to the matching degree value. For candidate result groups with the same matching degree, they are sorted in descending order according to the comprehensive score. For candidate result groups with the same comprehensive score, they are sorted in ascending order according to geographical distance. The geographical distance is calculated using the spherical distance formula between the user-defined geographical coordinates and the coordinates of the doctor's registered address. After sorting, the final display data list is generated. The front-end rendering engine displays the doctor's photo, name, qualification level, area of ​​expertise, location distance, comprehensive score, and matching degree value one by one according to the list order. The matching degree value is presented in the form of a progress bar.

[0029] The sixth step is the treatment plan generation and cost confirmation step. When the user clicks on any matching result to view detailed information, the treatment plan recommendation module and the cost calculation module are triggered. The treatment plan recommendation module selects the diagnostic opinion item that best matches the current symptoms of the user's pet based on the diagnostic opinion template library submitted by the physician. It then matches the corresponding acupuncture plan identifier with the constitution identification results. This acupuncture plan identifier points to a standardized acupoint combination scheme. The traditional Chinese medicine prescription is composed of corresponding classic prescriptions based on the type of illness. The massage techniques are selected based on the location of pain, and the dietary recommendations are selected based on the constitution type, along with corresponding food combinations. These four components are packaged to generate structured treatment plan data. The cost calculation module queries the local price list based on the service item code set in the treatment plan data to calculate various costs. The diagnosis fee is set as a one-time fixed amount, the treatment fee is accumulated based on the number of acupuncture and massage treatments, the traditional Chinese medicine fee is calculated per dose based on the prescription dosage, and the subsequent follow-up visit fee is estimated based on the number of follow-up visits. After summarizing, the total cost budget value for the user is obtained and displayed in the cost details area of ​​the front-end interface.

[0030] The seventh step is the payment and service loop. After reading the treatment plan and cost details, the user clicks the payment confirmation button. The payment module activates the third-party payment application interface installed on the user's mobile phone. After the user completes the payment authorization, the payment system returns a payment success status code. The platform records the transaction serial number of this matched service and updates the order status to paid. After the user completes the actual treatment service, the platform pushes a satisfaction evaluation questionnaire to invite the user to evaluate. The evaluation dimensions include three dimensions: satisfaction with treatment effect, satisfaction with service attitude, and satisfaction with cost reasonableness. The evaluation data submitted by the user is cleaned and written into the user evaluation table for subsequent comprehensive score calculation and dynamic adjustment and optimization of weight coefficients.

[0031] Since the scenario in this embodiment involves the diagnosis and treatment of arthritis in an elderly dog, the platform successfully matched a senior traditional Chinese veterinary veterinarian with more than ten years of experience, specializing in orthopedic diseases in elderly animals, and with a user rating of 4.8, according to the above workflow. The matching degree reached 0.89, and the user accepted the matching result and completed the subsequent diagnosis and treatment appointment and payment. Example 2

[0032] This embodiment is applied to a differentiated application scenario where pet owners seek out a traditional Chinese veterinary veterinarian skilled in diagnosing and treating skin conditions in their cats suffering from chronic skin diseases. In this scenario, the cat exhibits typical skin symptoms such as itching and hair loss. Pet owners are sensitive to treatment costs and desire veterinarians with experience in handling complex skin lesions. This embodiment is described below from two aspects: extended configuration of the system architecture and modified execution of the workflow.

[0033] In terms of system architecture, the data acquisition layer in this embodiment adds a functional depth to the TCM syndrome feature acquisition module based on the module configuration of Embodiment 1. Specifically, it expands to support automatic recognition and analysis of tongue images. The system integrates an image recognition model to extract color features and analyze texture features from user-uploaded tongue images, mapping the analysis results into standardized tongue image description data. This description data serves as a component of the syndrome vector in subsequent matching calculations. Simultaneously, the TCM veterinary resource database in this embodiment adds a physician dermatology specialist record table. This table records the number of cured skin lesions for each physician, their expertise in different skin disease syndrome types, and the number of characteristic TCM prescriptions used. The records in the dermatology specialist record table establish a query relationship with the basic information table through physician identifiers.

[0034] Regarding the weight configuration of the matching algorithm, this embodiment has made adaptive adjustments based on the characteristics of dermatology diagnosis and treatment. The adjusted weight coefficients are allocated as follows: pet physical characteristics 0.10, TCM syndrome characteristics 0.30, user preferences 0.15, disease type and severity 0.30, and doctor's qualifications and experience 0.15. Among these, the weight coefficients for TCM syndrome characteristics and disease type and severity are increased to enhance the matching accuracy of dermatology-specific features. The weighted matching calculation unit re-executes the similarity calculation according to the adjusted weight coefficients. During the calculation process, a comparison and matching step between tongue image analysis and skin lesion features is added. The tongue image analysis results are cross-matched with the dermatology syndrome classification of the doctor's expertise, and the conformity calculation adopts an exact matching mode.

[0035] In terms of workflow, the second step in this embodiment adds a tongue image intelligent analysis and processing step. After the user submits a tongue image, the system automatically calls the image recognition module for analysis, which takes about 2 to 3 seconds. After the analysis is completed, the analysis conclusion is automatically presented on the front end for the user to check and confirm. The logic of the remaining requirement parsing and vectorization conversion steps is consistent with that in Embodiment 1. After the matching calculation is completed, the system enters the sorting and display stage. Since the user selects a price range that is medium to low in the user preference settings, the sorting module adds a price filtering preprocessing step in this stage to filter out candidate doctors whose fees exceed the user-set price range. After filtering, the conventional three-level sorting process is then executed. The first item in the final matching result list is a chief physician of traditional Chinese veterinary medicine who specializes in the diagnosis and treatment of damp-heat syndrome skin diseases, whose fees are within the user-set range, and who has more than 50 historical cured cases.

[0036] This embodiment further demonstrates the actual operational effect of the dynamic weight adjustment mechanism. During continuous operation, the system monitors the matching satisfaction data for skin diseases and finds that the average matching satisfaction for skin diseases in the past 30 days is slightly lower than the preset threshold. The system automatically triggers the weight coefficient adjustment program, increasing the weight coefficient related to TCM syndrome characteristics by 0.02 based on the previous setting value, and correspondingly decreasing the weight coefficient related to user preferences by 0.02. The adjusted weight coefficients take effect immediately and are applied to the subsequent matching calculation formula. This dynamic adjustment mechanism ensures that the matching algorithm can continuously self-optimize and improve accuracy based on feedback from actual business data. Example 3

[0037] This embodiment applies to a backend management scenario where pet clinic operators import and centrally manage traditional Chinese veterinary medicine (TCM) medical resource data in bulk. In this scenario, management users need to register new TCM veterinary physicians in bulk, update existing physician qualification data, and adjust matching rule parameters. The following describes this embodiment from two aspects: the expansion of management function modules in the system architecture and the batch processing execution of workflows.

[0038] In terms of system architecture, this embodiment adds two management functional units to the business application layer: a resource batch import module and a rule parameter management module. The resource batch import module supports parsing and processing batch data files in CSV and Excel formats. Internally, the module integrates a data cleaning engine to perform format validation, null value detection, and duplicate checks on the imported data. The cleaning engine is configured with field mapping rules corresponding to the table structure of the Chinese veterinary resource database, and supports a customizable configuration interface for field correspondences. After importation, the system automatically generates an import report recording the number of successful and failed entries. The rule parameter management module provides a visual editing interface for the weight coefficients of filtering conditions. The interface displays the weight ratio values ​​of various filtering conditions in the form of a slider control. Administrators can adjust the weight values ​​by dragging the slider. The system calculates the sum of weights in real time and issues a warning when the sum deviates from 1.0. This module also supports setting thresholds for similarity calculation parameters and configuring and modifying the trigger threshold for dynamic adjustment.

[0039] In terms of workflow, the resource batch import process in this embodiment includes four steps. The first step is the file upload step, where the management user selects the prepared data file in the upload interface of the resource batch import module and confirms the upload. The file is transmitted to the server via a binary transmission protocol. The second step is the file parsing step, where the parsing engine reads the file content according to preset file format rules and converts each line of record into an intermediate data object. The third step is the data cleaning step, where the cleaning engine traverses all intermediate data objects to perform field integrity verification. Data objects with empty required fields are marked as verification failure records, and records that pass verification are checked for duplication. The check logic is to compare whether the physician's ID number and physician's license number are duplicates of existing records in the database. Duplicate records are marked as update operations. The fourth step is the database writing step, where the system writes each record to the corresponding database table according to the writing mark. After the writing is completed, a formatted report is generated to display the import result statistics.

[0040] The rule parameter management process includes three steps: parameter viewing, parameter editing, and parameter saving. Users can view the currently effective weight coefficients and threshold configurations on the main interface of the parameter management module. After clicking the edit button to enter edit mode, they can adjust the values ​​of each parameter using the slider control. The system displays the changes in the sum of weights in real time. When the sum of weights equals 1.0, the save button becomes active. After the user clicks the save button, the system writes the new parameters to the cached key-value pairs in the matching rule database and makes them effective immediately. Hot updates of parameters can be achieved without restarting the service platform.

[0041] This embodiment also demonstrates the operation mechanism of the matching closed-loop verification module. After the user completes the diagnosis and treatment service, the platform automatically pushes a satisfaction evaluation questionnaire. The questionnaire is sent to the user's mobile terminal through an instant messaging channel. The questionnaire includes three scoring items: evaluation of treatment effect, evaluation of service attitude, and evaluation of cost reasonableness, as well as a text opinion input box. The questionnaire data submitted by the user is encrypted and transmitted back to the platform and written into the corresponding field of the user evaluation table. At the same time, the weight adjustment analysis program is triggered to perform statistical analysis on historical evaluation data, providing data support for the next round of dynamic weight adjustment, thus forming a complete matching service closed-loop verification process.

[0042] As can be seen from the detailed description of the three embodiments above, the pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening involved in this invention, through a systematic multi-level architecture design and a procedural matching algorithm, can effectively achieve accurate matching of multi-dimensional attributes such as pet physical signs, TCM syndrome characteristics, user preferences and disease types with TCM veterinary diagnosis and treatment resources. Furthermore, the matching algorithm is continuously optimized through a dynamic weight adjustment mechanism, which effectively improves the efficiency and accuracy of pet owners in finding suitable TCM veterinary doctors.

Claims

1. A pet TCM diagnosis and treatment service platform and matching method based on multi-condition screening, characterized in that, Includes the following steps: We standardize the collection of multidimensional attribute data on pet vital signs, TCM syndrome characteristics, user preferences, and disease types and severity to construct a multidimensional attribute set for pet medical needs. We standardize the mapping of TCM veterinary physicians' professional direction, professional qualifications, syndrome types of expertise, clinical experience, fee standards, and user reviews to establish a matching rule base. We use weighted matching algorithms or similarity models to match the demand attribute vector with the TCM veterinary resource node attribute vector to generate a candidate matching result list. We sort and display the candidate matching results according to matching degree, comprehensive score, and distance. We collect historical service feedback and satisfaction evaluation data and dynamically adjust the weight coefficients to continuously optimize the matching algorithm.

2. The method as described in claim 1, characterized in that, The pet's physical characteristics information in step S1 includes breed, age, weight, sex, and past medical history; the TCM syndrome characteristics include the results of constitution identification, tongue appearance, pulse appearance, and coat color; the user preferences include geographical location, price range, appointment time, and doctor qualification level; the disease type and severity include the specific disease name and the stage of disease development.

3. The method as described in claim 1, characterized in that, The standardized mapping described in step S2 includes: a mapping between the doctor's professional direction and common pet disease types; a mapping between the level of professional qualification and the level of the certification body; a mapping between the syndrome types of expertise and the TCM syndrome differentiation system; a mapping between years of clinical experience and the number of successful cases; a mapping between fee standards and service items; and a mapping between user reviews and service satisfaction scores.

4. The method as described in claim 1, characterized in that, The weight coefficients of the weighted matching algorithm described in step S3 are set according to the importance of pet physical characteristics, TCM syndrome characteristics, user preferences, disease type and severity, and doctor's qualifications and experience. The sum of each weight coefficient satisfies the preset normalization condition.

5. The method as described in claim 1, characterized in that, The similarity model described in step S3 uses the cosine similarity algorithm, which takes the cosine of the angle between the demand attribute vector and the attribute vector of each veterinary resource node as the similarity value.

6. The method as described in claim 1, characterized in that, The priority rules for sorting in step S4 are as follows: first, sort by matching degree in descending order; when the matching degree is the same, sort by overall score in descending order; when the overall score is the same, sort by distance in ascending order.

7. The method as described in claim 1, characterized in that, The triggering conditions for dynamically adjusting the weight coefficients in step S5 include: when the matching satisfaction of a certain type of disease is continuously lower than a preset threshold within a preset time period, the weight coefficient of the relevant attributes of that type of disease is automatically increased; when a doctor's historical positive review rate rises to a new level, the weight coefficient of the doctor's qualification attributes is automatically increased.

8. The method as described in claim 1, characterized in that, It also includes a treatment plan recommendation module, which automatically generates personalized treatment plans based on the diagnosis opinions of traditional Chinese veterinary physicians in the matching results, combined with the pet's physical condition identification results and current syndrome characteristics. These plans include acupuncture plans, traditional Chinese medicine prescriptions, massage techniques, and dietary therapy suggestions.

9. The method as described in claim 1, characterized in that, It also includes a cost calculation and payment integration module, which automatically calculates the estimated costs of diagnosis, treatment, medication, and follow-up visits based on the recommended treatment plan, and integrates with third-party payment platforms to enable online payment.

10. The method as described in claim 1, characterized in that, It also includes a matching closed-loop verification module, which pushes a satisfaction evaluation questionnaire after the user completes the diagnosis and treatment service, collects evaluation data on dimensions such as treatment effect, service attitude and cost reasonableness, and stores the evaluation results in the database for weight coefficient optimization.

Images