User portrait analysis on Chinese diabetes online health platforms: an information processing perspective

Purpose The objective of this research is to investigate the characteristics of information interaction among users of the largest online health platform for diabetes in China from an information processing viewpoint, determine the stages of information processing among users and reveal the variations in information requirements and behavioral patterns across different user groups at various processing levels, ultimately creating a user segmentation labeling system to enhance user portrait. Design/methodology/approach This study adopts a deep learning BILSTM-CNN classification model to identify user information processing characteristics, and then classify users into three groups. The LDA topic model is employed to analyze the information needs of these groups. Findings This research utilizes a BILSTM-CNN combined deep learning model, showcasing enhanced effectiveness in classifying the degree of information processing in user comments. Our model also increases the accuracy and stability of classification compared to conventional deep learning models, achieving an F1 score of 95.0% (F1 Score: CNN 92%, LSTM 94%, BILSTM 94%). Based on the classification results, users were grouped and different groups of diabetes users showed differences in information needs, information behavior and natural attributes. Originality/value Taking information processing as an entry point, this study deeply mined the user behavior data of China’s largest diabetes online health platform, identifying the information processing characteristics present in user comments and categorizing users into groups reflecting varying depths of information processing. Based on the multi-dimensional data analysis, we innovatively constructed a refined user labeling system, and finally depicted a complete user portrait. This study not only enriches the theoretical framework of cognitive processing and user portrait but also contributes to research on personalized recommendations for online health platforms for diabetes. Peer review The peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-11-2024-0728.