Multi-source health data processing method and device, and electronic device

Abstract

The application relates to the technical field of data processing and health management, and provides a multi-source health data processing method, device and electronic equipment. The multi-source health data processing method comprises the following steps: acquiring health data input by a plurality of intelligent health devices; performing quality detection and cleaning on the health data; performing conflict identification on the cleaned health data, and retaining optimal health data from the conflict identification result according to a preset conflict resolution strategy; converting the optimal health data into a preset standard format for storage, and generating a migration report, wherein the migration report comprises at least one of a data cleaning log, a conflict resolution log and a data conversion log; the conflict problem of multi-source data can be intelligently solved, a continuous and complete health record can be generated, data barriers can be broken, health data can be seamlessly migrated across brands, and therefore the user experience is improved.

Description

Technical Field

[0001] This application relates to the field of data processing and health management technology, specifically to a method, apparatus, and electronic device for processing multi-source health data. Background Technology

[0002] Currently, there are various brands of smart health devices on the market (such as Xiaomi body fat scale, Huawei smart scale, Youpin scale, etc.), and each brand provides a matching health management application to record and display users' health data (such as weight, BMI, body fat percentage, etc.). As users' awareness of health management increases, many users switch devices between different brands or use multiple devices interchangeably, resulting in personal health data being scattered across multiple isolated data files.

[0003] To address the aforementioned issue of data fragmentation, relevant technologies primarily offer two solutions: Solution 1 is single-brand data silo management, where users can only view data within the brand's own health management application, and the data is account-bound and cannot be separated from its original ecosystem; Solution 2 is semi-automatic import based on file format, where some health management applications support manual import of CSV / Excel files, but require users to strictly adhere to the application's internal format for data preparation. During import, users need to manually select the file type, specify the brand source, and configure field correspondences one by one. Although these solutions provide the possibility of data export and import, they cannot intelligently resolve conflicts between multiple data sources when handling multi-source data fusion, resulting in the inability to generate continuous and complete health records, thus impacting user experience. Summary of the Invention

[0004] In view of the above problems, embodiments of this application provide a multi-source health data processing method, apparatus and electronic device to solve the above technical problems.

[0005] In a first aspect, embodiments of this application provide a method for processing multi-source health data, comprising: acquiring health data input from multiple smart health devices; performing quality inspection and cleaning on the health data; identifying conflicts in the cleaned health data and retaining the optimal health data from the conflict identification results according to a preset conflict resolution strategy; converting the optimal health data into a preset standard format for storage and generating a migration report, wherein the migration report includes at least one of a data cleaning log, a conflict resolution log, and a data conversion log; capable of intelligently resolving conflicts in multi-source data, generating continuous and complete health records, breaking down data barriers, and achieving seamless cross-brand migration of health data, thereby improving user experience.

[0006] Secondly, embodiments of this application also provide a multi-source health data processing device, comprising: The data acquisition module is used to acquire health data input from multiple smart health devices; The quality inspection and cleaning module is used to perform quality inspection and cleaning on the health data; The conflict identification and resolution module is used to identify conflicts in the cleaned health data and retain the optimal health data from the conflict identification results according to a preset conflict resolution strategy. The generation module is used to convert the optimal health data into a preset standard format for storage and generate a migration report. The migration report includes at least one of data cleaning logs, conflict resolution logs, and data transformation logs. It can intelligently resolve conflicts between data from multiple sources, generate continuous and complete health records, break down data barriers, and achieve seamless cross-brand migration of health data, thereby improving user experience.

[0007] Thirdly, this application also provides an electronic device, including a device body and the multi-source health data processing device disposed on the device body; it can intelligently resolve the conflict problem of multi-source data, generate continuous and complete health records, break down data barriers, and realize seamless migration of health data across brands, thereby improving user experience.

[0008] This application provides a multi-source health data processing method, comprising: acquiring health data input from multiple smart health devices; performing quality inspection and cleaning on the health data; identifying conflicts in the cleaned health data and retaining the optimal health data from the conflict identification results according to a preset conflict resolution strategy; converting the optimal health data into a preset standard format for storage and generating a migration report, wherein the migration report includes at least one of a data cleaning log, a conflict resolution log, and a data conversion log; intelligently solving problems in the prior art such as difficulty in conflict resolution, poor data continuity, and cross-device migration gaps caused by heterogeneous multi-source data formats, redundancy, and manual intervention, generating continuous and complete health records, breaking down data barriers, and achieving seamless cross-brand migration of health data, thereby improving user experience.

[0009] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This application illustrates a multi-source health data processing system provided in an embodiment of the present application.

[0012] Figure 2 This application illustrates a multi-source health data processing apparatus provided in an embodiment of the present application.

[0013] Figure 3 This application illustrates a method for processing multi-source health data according to an embodiment of the present application.

[0014] Figure 4 This application illustrates a method for processing multi-source health data according to another embodiment.

[0015] Figure 5 This application illustrates a method for processing multi-source health data according to another embodiment.

[0016] Figure 6 This application illustrates a method for processing multi-source health data according to another embodiment.

[0017] Figure 7 This application illustrates a method for processing multi-source health data according to another embodiment. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of the embodiments are shown 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 only used to explain this application, and should not be construed as limiting this application.

[0019] To enable those skilled in the art to better understand the solutions of this application, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0020] In the embodiments of this application, it should be noted that, in this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0021] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0022] In the description of the embodiments of this application, the words "example" or "for example" are used to indicate exemplification, illustration, or description. Any embodiment or design described as "example" or "for example" in the embodiments of this application is not to be construed as being more preferred or having more advantages than another embodiment or design. The use of the words "example" or "for example" is intended to present relative concepts in a clear manner.

[0023] Furthermore, in the embodiments of this application, "multiple" refers to two or more. Therefore, in the embodiments of this application, "multiple" can also be understood as "at least two". "At least one" can be understood as one or more, such as one, two, or more. For example, including at least one means including one, two, or more, and is not limited to which ones are included. For example, including at least one of A, B, and C, then it could include A, B, C, A and B, A and C, B and C, or A and B and C.

[0024] It should be noted that in the embodiments of this application, "and / or" describes the relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. In addition, the character " / ", unless otherwise specified, generally indicates that the associated objects before and after it are in an "or" relationship.

[0025] As users become more health-conscious, many switch devices across brands or use multiple devices to manage health-related data, resulting in personal health data being scattered across multiple isolated data files. Current solutions typically restrict users to viewing and managing data only within the brand's own health management application, with some applications supporting manual import of CSV files.

[0026] In one application scenario, a user initially recorded their weight data for a year using a Xiaomi body fat scale, and later switched to a smart scale from the Huawei or Apple ecosystem. The user wants to see their complete weight trend over the past year within the Huawei / Apple health management app, rather than starting the recording process all over again. In this application, the user uploads a CSV file exported from Xiaomi to this device. After data cleaning, conflict identification and resolution, and conversion to a standard format compatible with Huawei / Apple health management apps, the weight data is seamlessly migrated. Users do not need to manually select file formats or specify field mappings, thus improving the user experience.

[0027] like Figure 1 As shown, Figure 1 A multi-source health data processing system 100 provided in an embodiment of this application is illustrated schematically. The multi-source health data processing system 100 includes one or more smart health devices 10 and a multi-source health data processing device 20. The smart health devices 10 and the multi-source health data processing device 20 are connected via wired (e.g., Wi-Fi / 4G / 5G / NFC) or wireless (e.g., USB / BLE) means. The smart health devices 10 are smart health data collection terminals covering various types and brands, including but not limited to Xiaomi body fat scales, Huawei smart scales, and Youpin scales. Figure 2 As shown, the multi-source health data processing device 20 includes a data acquisition module 21, a quality inspection and cleaning module 22, a conflict identification and resolution module 23, and a generation module 24. The data acquisition module 21 acquires health data input from multiple smart health devices 10; the quality inspection and cleaning module 22 performs quality inspection and cleaning on the health data; the conflict identification and resolution module 23 identifies conflicts in the cleaned health data and retains the optimal health data from the conflict identification results according to a preset conflict resolution strategy; the generation module 24 converts the optimal health data into a preset standard format for storage and generates a migration report, which includes at least one of a data cleaning log, a conflict resolution log, and a data conversion log. This multi-source health data processing system 100 intelligently solves problems in existing technologies caused by heterogeneous multi-source data formats, redundancy, and manual intervention, such as difficulty in conflict resolution, poor data continuity, and cross-device migration gaps. It generates continuous and complete health records, breaks down data barriers, and enables seamless cross-brand migration of health data, thereby improving user experience.

[0028] like Figure 3 As shown, Figure 3 The multi-source health data processing method provided in this application embodiment is illustrated schematically. The method includes the following steps: Step S10: Obtain health data input from multiple smart health devices.

[0029] In this step, smart health devices include, but are not limited to, Xiaomi body fat scales, Huawei smart scales, and Youpin scales. Health data includes at least one of the following: weight, BMI, body fat percentage, and time. The health data input into the smart health devices can be in CSV, Excel, or JSON format.

[0030] Step S20: Perform quality checks and cleaning on the health data.

[0031] This step involves quality inspection and cleaning, including outlier detection, duplicate data detection, and integrity checks. Outlier detection includes one or more of the following: abnormal weight, abnormal BMI, abnormal body fat percentage, and abnormal time. Abnormal weight is exemplified by a weight outside the 20kg-300kg range. Abnormal BMI is exemplified by a BMI outside the 10-60 range. Abnormal body fat percentage is exemplified by a body fat percentage outside the 5%-50% range. Abnormal time is exemplified by a measurement time earlier than 2000-01-01 or later than the current system time plus 24 hours. Once abnormal data is detected, it is marked as "abnormal" and excluded from statistical analysis. Automatic filtering of abnormal data improves data reliability. Duplicate data detection uses a data fingerprint comparison mechanism. For example, it calculates the MD5 hash of the data (time + weight) and compares the MD5 fingerprints of two data points. If the MD5 fingerprints of two data points are identical, they are considered completely duplicated. The earliest record is retained, and the remaining duplicates are deleted. This method quickly identifies duplicates, intelligently removes duplicates, and avoids statistical errors caused by data redundancy. Integrity checks detect whether key fields are missing. Key fields include weight, BMI, body fat percentage, and time. For example, if a data record is missing the "weight" or "time" field, it is marked as "incomplete data" and import is prohibited. If the "time" field is missing but other fields (such as weight, BMI, and body fat percentage) exist, the last modified time of the data file is used as the default time, with an "estimated time source" label added. If only fields such as BMI and body fat percentage are missing, the basic data is imported normally, and the fields are set to null values. Through data integrity checks, it is ensured that no key fields are missing.

[0032] Step S30: Conflict identification is performed on the cleaned health data, and the best health data is retained from the conflict identification results according to the preset conflict resolution strategy.

[0033] In this step, conflict identification includes temporal proximity conflicts and numerical difference conflicts. Temporal proximity conflicts target redundant data that are from the same source or are exactly the same; numerical difference conflicts target contradictory data that are from different sources and have similar values. For example, a user weighed themselves once in the morning using a Xiaomi scale (70.5kg), and then again five minutes later using a Huawei scale (71.8kg). Neither of these are incorrect data, but they cannot be directly merged; the optimal health data needs to be retained from the two.

[0034] The preset conflict resolution strategies include at least one of the following: a data integrity priority strategy, a device credibility score priority strategy, a time accuracy priority strategy, and a user operation strategy. Specifically, the data integrity priority strategy involves comparing the number of fields in each conflict identification result and retaining the health data with the most fields; the device credibility score priority strategy involves querying the preset device credibility score based on the data source brand of each conflict identification result and retaining the health data with the highest score; and the time accuracy priority strategy involves comparing the time accuracy of each conflict identification result and retaining the health data with the highest accuracy. The user operation strategy refers to displaying two data comparisons in a pop-up window when the data integrity priority strategy, device credibility score priority strategy, and time accuracy priority strategy cannot determine which data to retain, requesting the user to manually select which data to keep.

[0035] Step S40: Convert the optimal health data into a preset standard format for storage and generate a migration report. The migration report includes at least one of the following: data cleaning log, conflict resolution log, and data transformation log.

[0036] In this step, when the optimal health data is finally imported into the database or written to a file, a preset standard format is used, such as FHIR (Fast Healthcare Interoperability Resources) or a custom flat JSON structure. Weight is uniformly expressed in kg, and time is uniformly formatted as ISO in UTC+8 time zone. Finally, a migration report in PDF or HTML format is generated. The migration report includes at least one of the following: data cleaning log, conflict resolution log, and data transformation log. For example: Data source: Xiaomi body fat scale; file format: CSV; import time: 2025-01-20 10:00:00; total records: 180; successful import: 173; abnormal data: 2 (isolated); duplicate data: 5 (deduplicated), etc.

[0037] The multi-source health data processing method of this application embodiment identifies conflicts in health data and retains the optimal health data from the conflict identification results according to a preset conflict resolution strategy. It can intelligently solve the problems in the prior art caused by heterogeneous multi-source data formats, redundancy, and manual intervention, such as difficulty in conflict resolution, poor data continuity, and cross-device migration gaps. It generates continuous and complete health records, breaks down data barriers, and realizes seamless cross-brand migration of health data, thereby improving user experience.

[0038] As one embodiment, after step S40, the method further includes: providing a data preview interface to display the first 10 converted data entries for user confirmation; after user confirmation, storing standardized data in the target database and adding a traceability identifier to each data entry so that users can view the source information of each data entry in the system at any time (such as brand, device model, import time, original format, etc.), further improving the user experience.

[0039] As one example, please refer to Figure 4 Step S30 further includes the following steps: Step S301: Calculate the time difference and / or weight difference between various health data.

[0040] In this step, as one embodiment, only the time difference between each health data point is calculated. As another embodiment, only the weight difference between each health data point is calculated. As yet another embodiment, both the time difference and weight difference between each health data point are calculated simultaneously.

[0041] Step S302: When the time difference is lower than the first preset threshold and / or the weight difference is lower than the second preset threshold, retain the optimal health data from the conflict identification results according to the preset conflict resolution strategy.

[0042] In this step, the first and second preset thresholds can be set by the user or by the system based on the application scenario and historical data. For example, the first preset threshold is 5 minutes, and the second preset threshold is 0.5 kg. For instance, the time difference between various health data points is calculated. When the time difference is less than or equal to 5 minutes, the optimal health data is retained from the conflict identification results according to a preset conflict resolution strategy. Similarly, the weight difference between various health data points is calculated. When the weight difference is less than or equal to 0.5 kg, the optimal health data is retained from the conflict identification results according to a preset conflict resolution strategy. Finally, the time difference and weight difference between various health data points are calculated. When the time difference is less than or equal to 5 minutes and the weight difference is less than or equal to 0.5 kg, the optimal health data is retained from the conflict identification results according to a preset conflict resolution strategy.

[0043] As one embodiment, retaining the optimal health data from the conflict identification results according to the preset conflict resolution strategy specifically involves sequentially executing the data integrity priority strategy, the device credibility score priority strategy, the time accuracy priority strategy, and the user operation strategy to retain the optimal health data from the conflict identification results.

[0044] For example, if the conflict identification result includes data A and data B, where data A contains: weight + time + BMI + body fat percentage; and data B contains: weight + time; then when implementing the data integrity priority strategy, data A is retained.

[0045] For example, suppose the preset device credibility scores are: 95 points for medical-grade devices (such as InBody), 85 points for professional body fat scales (such as Huawei / Apple ecosystem devices), 85 points for ordinary electronic scales (such as Xiaomi / Youpin), and 60 points for mechanical scales. If the conflict identification result includes data A and data B, where data A comes from InBody and data B comes from a Xiaomi scale, then the device credibility score priority strategy is executed, and data A is retained.

[0046] For example, if the conflict identification result includes data A and data B, where the time of data A is 2025-04-05 08:30:10 and the time of data B is 2025-04-05 08:30, then when the time accuracy priority strategy is executed, the data accurate to the second level, i.e., data A, is retained.

[0047] When the above automatic strategy fails (e.g., both Huawei scales score 85), a modal dialog box pops up, displaying key differences (such as numerical value, source, and time). The system-recommended option (usually the one with the higher score) is selected by default, and the user clicks "Confirm" to complete the selection.

[0048] As one example, please refer to Figure 5 The following steps are included after step S10 and before step S20: Step S50: Perform format checks on the health data.

[0049] In this step, the file extension is read, and the file format type is determined based on the extension. For example, if the extension is .csv, it is identified as CSV format; if the extension is .xlsx or .xls, it is identified as Excel format; if the extension is .json, it is identified as JSON format; and if the extension is .txt, it is identified as CSV format.

[0050] Step S60: Automatically identify the brand source of the health data based on the test results and generate brand classification results.

[0051] In this step, brand classification can be performed using field matching or machine learning models.

[0052] As one example, brand classification is performed using field matching; please refer to [link to relevant documentation]. Figure 6 Step S60 includes the following steps: Step S601: Construct a brand feature fingerprint database. The brand feature fingerprint database shall include at least the data structure features, feature fields, time format, and unit format of each brand.

[0053] In this step, data structure features such as column order, delimiters, and nesting levels are considered. For example, Xiaomi's CSV format feature fields include: "weight (kg)", "body fat percentage (%)", and "measurement time"; Huawei's Excel format feature fields include: "weight_kg", "bmi", and "timestamp"; and Youpin's JSON format feature fields include: "bodyWeight": {"value": 70, "unit": "kg"}, "time": "2024-05-01T08:30:00+08:00".

[0054] Step S602: Extract the field names from the health data and match the field names with the brand feature fingerprint database.

[0055] In this step, the file header is first parsed according to the file format to extract the field names. For example, the first row of a CSV file is read, the first row of the first sheet of an Excel file is read, and the key name of the top-level object is extracted from a JSON file. Then, the feature fields in the brand feature fingerprint database are traversed to obtain the matching feature fields.

[0056] Step S603: Calculate the matching degree between health data and each brand, and select the brand with the highest matching degree as the brand classification result.

[0057] In this step, the matching degree is the ratio of the number of matching feature fields to the total number of field names in the health data. The brand with the highest matching degree is selected as the brand classification result. By calculating the matching degree of feature fields, the source of health data can be accurately identified.

[0058] As one example, a machine learning model is used for brand classification; please refer to [link / reference]. Figure 7 Step S60 includes the following steps: Step S604: Extract data features from the detection results. The data features include at least: file structure features, feature fields, time format, and unit format.

[0059] Step S605: Input the data features into the pre-trained machine learning model to classify brands and output the corresponding brand probability distribution.

[0060] In this step, machine learning models such as random forests or lightweight neural networks take data features as input and output the corresponding brand probability distribution.

[0061] Step S606: Select the brand with the highest probability as the brand classification result.

[0062] Step S70: Load the corresponding preset field mapping rules according to the brand classification results, and extract the corresponding field values ​​from the health data according to the preset field mapping rules.

[0063] In this step, preset field mapping rules map the field names of each brand to unified standard field names, establishing a "many-to-one" mapping relationship. Preset field mapping rules can be stored in a configuration file; when adding a new brand, only the configuration file needs to be updated, without modifying the code. For example, preset field mapping rules are shown in Table 1: Iterate through each row of health data and extract field values ​​according to preset field mapping rules. For example, Xiaomi's health data is 62.5, 1705286400, 21.8, 18.5. After mapping: weight = 62.5, time = 1705286400 (to be converted), BMI = 21.8, body fat percentage = 18.5.

[0064] Step S80: Perform data standardization processing based on the extraction results.

[0065] In this step, data standardization includes time format standardization, unit standardization, and numerical precision standardization. Time format standardization can be uniformly converted to ISO 8601 format, such as 2024-01-15T07:30:00+08:00; unit standardization can be converting units such as "pound" and "jin" to "kilogram"; numerical precision standardization can be adjusting the number of decimal places, such as retaining one decimal place for weight, two decimal places for BMI, and one decimal place for body fat percentage.

[0066] In the application scenario of importing data from the Xiaomi Body Fat Scale, for example, assuming a user uses the Xiaomi Body Fat Scale for 180 days, the health data imported from the Xiaomi Body Fat Scale is exported as "mi_health.csv": weight, timestamp, BMI, body fat 62.5, 1705286400, 21.8, 18.5 62.3, 1705372800, 21.7, 18.4 The specific processing flow is as follows: 1. Identify the format: CSV; 2. Extract fields: weight, timestamp, bmi, body_fat; 3. Match the brand: Xiaomi (90% matching rate); 4. Field mapping: weight → body weight, timestamp → time, bmi → BMI, body_fat → body fat percentage; 5. Data standardization: convert Unix timestamps to ISO 8601 format; 6. Quality inspection and cleaning: 180 data entries, 2 abnormal entries, 5 duplicate entries; 7. Successfully imported after conflict resolution: 173 valid data entries.

[0067] This embodiment addresses the issues of automatic brand recognition and format incompatibility through feature field matching, and achieves flexibility and scalability in field mapping through dynamic field mapping rules. Users can complete data migration with a single click without manual configuration, avoiding tedious operations and improving data migration efficiency, significantly enhancing the user experience.

[0068] This application also provides an electronic device, which includes a device body and a multi-source health data processing device as described above, disposed within the device body. The electronic device may be, but is not limited to, a weight scale, body fat scale, nutrition scale, smart wearable device, or mobile terminal. Smart wearable devices include, but are not limited to, smartwatches and smart bracelets. Mobile terminals include, but are not limited to, smartphones, laptops, and tablets. This electronic device acquires health data input from multiple smart health devices; performs quality inspection and cleaning on the health data; identifies conflicts in the cleaned health data and retains the optimal health data from the conflict identification results according to a preset conflict resolution strategy; converts the optimal health data into a preset standard format for storage; and generates a migration report containing data cleaning logs, conflict resolution logs, and data conversion logs. It can intelligently solve problems in existing technologies such as difficulty in conflict resolution, poor data continuity, and cross-device migration gaps caused by heterogeneous multi-source data formats, redundancy, and manual intervention, generating continuous and complete health records, breaking down data barriers, and achieving seamless cross-brand migration of health data, thereby improving user experience.

[0069] The above are merely preferred embodiments of this application and are not intended to limit this application in any way. Although this application has disclosed preferred embodiments as above, it is not intended to limit this application. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the technical solution of this application. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of this application without departing from the scope of the technical solution of this application shall still fall within the scope of the technical solution of this application.

Claims

1. A method for processing multi-source health data, characterized in that, include: Acquire health data input from multiple smart health devices; The health data is subjected to quality testing and cleaning. The cleaned health data is subjected to conflict identification, and the optimal health data is retained from the conflict identification results according to the preset conflict resolution strategy; The optimal health data is converted into a preset standard format for storage, and a migration report is generated. The migration report includes at least one of the following: data cleaning log, conflict resolution log, and data transformation log.

2. The multi-source health data processing method as described in claim 1, characterized in that, The health data includes at least one of: weight, BMI, body fat percentage, and time; the step of conflict identification of the cleaned health data and retaining the optimal health data from the conflict identification results according to a preset conflict resolution strategy includes: Calculate the time difference and / or weight difference between the various health data points; When the time difference is lower than a first preset threshold and / or the weight difference is lower than a second preset threshold, the optimal health data is retained from the conflict identification results according to the preset conflict resolution strategy.

3. The multi-source health data processing method as described in claim 2, characterized in that, The preset conflict resolution strategy includes at least one of the following: data integrity priority strategy, device reliability score priority strategy, time accuracy priority strategy, and user operation strategy; The data integrity priority strategy includes: comparing the number of fields in each conflict identification result and retaining the healthy data with the most fields; The device credibility scoring priority strategy includes: querying the preset device credibility score based on the data source brand of each conflict identification result, and retaining the health data with the highest score; The time accuracy priority strategy includes: comparing the time accuracy of each conflict identification result and retaining the health data with the highest accuracy.

4. The multi-source health data processing method as described in claim 3, characterized in that, When the time difference is lower than a first preset threshold and / or the weight difference is lower than a second preset threshold, retaining the optimal health data from the conflict identification results according to a preset conflict resolution strategy includes: When the time difference is lower than the first preset threshold and / or the weight difference is lower than the second preset threshold, the data integrity priority strategy, the device credibility score priority strategy, the time accuracy priority strategy, and the user operation strategy are executed in sequence to retain the optimal health data from the conflict identification results.

5. The multi-source health data processing method as described in claim 1, characterized in that, The quality inspection and cleaning of the health data includes: The health data is then subjected to outlier detection, duplicate data detection, and integrity detection in sequence.

6. The multi-source health data processing method as described in claim 1, characterized in that, After acquiring health data input from multiple smart health devices and before performing quality detection and cleaning on the health data, the process also includes: The health data is then subjected to format detection. The system automatically identifies the brand source of the health data based on the test results and generates brand classification results. Based on the brand classification results, load the corresponding preset field mapping rules, and extract the corresponding field values ​​from the health data according to the preset field mapping rules; The extracted results are then subjected to data standardization processing.

7. The multi-source health data processing method as described in claim 6, characterized in that, The step of automatically identifying the brand source of the health data based on the test results and generating brand classification results includes: Construct a brand feature fingerprint database, which includes at least: the data structure features, feature fields, time format, and unit format of each brand; Extract the field names from the health data and match the field names with the brand feature fingerprint database; Calculate the matching degree between the health data and each brand, and select the brand with the highest matching degree as the brand classification result.

8. The multi-source health data processing method as described in claim 6, characterized in that, The step of automatically identifying the brand source of the health data based on the detection results and generating brand classification results includes: Data features are extracted from the detection results, and the data features include at least: file structure features, feature fields, time format, and unit format; The data features are input into a pre-trained machine learning model for brand classification, and the corresponding brand probability distribution is output. The brand with the highest probability is selected as the brand classification result.

9. A multi-source health data processing device, characterized in that, include: The data acquisition module is used to acquire health data input from multiple smart health devices; The quality inspection and cleaning module is used to perform quality inspection and cleaning on the health data; The conflict identification and resolution module is used to identify conflicts in the cleaned health data and retain the optimal health data from the conflict identification results according to a preset conflict resolution strategy. The generation module is used to convert the optimal health data into a preset standard format for storage and generate a migration report, which includes at least one of a data cleaning log, a conflict resolution log, and a data transformation log.

10. An electronic device, characterized in that, It includes a main body of the device and a multi-source health data processing device as described in claim 9, which is disposed on the main body of the device.

Images