体温测量数据集

1、数据采集：通过医卫通AI一体机智能建档，在用户授权下依法合规采集体温测量数据。原始数据包括唯一设备标识（健康设备ID）、系统生成的匿名用户标识（健康设备监测用户ID）、体温数值（单位：℃）、测量时间戳（即“体温采集时间”）及系统自动生成的状态标记（处理异常标识）。
2、数据处理：1）以“健康设备ID”和“体温采集时间”为联合主键，对原始数据进行清洗，剔除异常值（如<35℃或>42℃）；2）计算多维衍生指标：个体维度——日均体温、发热事件频次（≥37.3℃）；设备维度——日均服务人数、高峰使用时段、设备利用率；群体维度——区域发热率、异常体温事件密度；3）生成结构化、标准化的体温测量数据集，支持个体分析、设备管理与群体预警。
3、数据应用：输出高质量的多用户体温数据，用于生成个人体温趋势报告、触发高体温预警、评估设备服务效能、优化社区健康服务资源配置及训练用户切换识别与传染病传播预测AI模型。

1. Data Collection: Body temperature measurement data is collected in compliance with applicable laws and regulations with explicit user authorization via the Yiwitong AI All-in-One Device for intelligent health record creation. The raw data includes unique device identifier (Health Device ID), system-generated anonymous user identifier (Health Device Monitoring User ID), body temperature value (unit: ℃), measurement timestamp (i.e., "body temperature collection time"), and system-generated status marker (processing abnormality flag).
2. Data Processing: 1) Clean the raw data with "Health Device ID" and "body temperature collection time" as the composite primary key, and remove outliers (e.g., values <35℃ or >42℃); 2) Calculate multi-dimensional derived metrics: Individual-level: daily average body temperature, frequency of fever events (body temperature ≥37.3℃); Device-level: daily average number of served users, peak usage hours, device utilization rate; Population-level: regional fever rate, density of abnormal body temperature events; 3) Generate a structured and standardized body temperature measurement dataset that supports individual analysis, device management and population-level early warning.
3. Data Application: High-quality multi-user body temperature data is output, which is used to generate individual body temperature trend reports, trigger high-body-temperature early warnings, evaluate device service effectiveness, optimize the allocation of community health service resources, and train AI models for user switching recognition and infectious disease transmission prediction.

搜集汇总

背景与挑战

背景概述

该体温测量数据集规模为15KB，每月更新，来源于多来源，属于卫生行业。它记录非专属性场景下的体温数据，支持群体健康监测、发热筛查和设备效率评估，并通过匿名ID实现隐私保护下的多用户建模。数据集经过清洗和衍生指标计算，用于AI模型训练和公共卫生应急响应。