Abstract: With the growing prevalence of internetbased medical services, the value mining of online healthcare and medical derivative data has become increasingly significant. Developing a scientifically grounded evaluation index system facilitates the accurate assessment and effective utilization of data value. We adopted grounded theory to construct the evaluation framework. Through a systematic review of academic literature, three levels of coding were conducted to construct an initial indicator framework. Based on this framework, expert interviews were used to revise and refine the applicability of the indicators. The final evaluation system comprises three primary dimensions—intrinsic value, extrinsic value, and cost—along with 8 secondary indicators and 37 tertiary indicators. The Delphi method was employed to gather input from 12 experts, and the Analytic Hierarchy Process(AHP) was applied to determine the initial indicator weights. Subsequently, a Backpropagation(BP) neural network model was developed. Using 300 user samples, the model was trained via error backpropagation, resulting in a datadriven optimization of the indicator weights. The findings indicate that intrinsic value constitutes the core of healthcarederived data assessment, with data authenticity, completeness, and timeliness being particularly critical. The analysis of extrinsic value highlighted the broad impact of data in realworld applications. In the cost dimension, the significance of privacy protection and risk response increased notably after optimization. The proposed evaluation system offers a practical tool for assessing the value of online healthcare and medical derivative data and provides actionable insights for platform operators seeking to enhance their data management strategies.

Key words: Internet healthcare and medical services, Derivative data, Value evaluation, Grounded theory