关键词: 跨平台, 知识交流行为, 跨平台用户分类, CRFM模型, K-means

Abstract: [Purpose/Significance] With the diversification of online platforms, users tend to acquire and share knowledge across multiple platforms. Therefore, focusing on the classification of cross-platform users is of significance for accurately identifying cross-platform users, understanding their cross-platform knowledge exchange behaviors, and uncovering the structure of cross-platform behavior system. [Design/Methodology] In this paper, we take 120 users in Bilibili knowledge board as the research object. Based on users alignment, we obtain their attributes, content creation , and interaction data both on Bilibili and Weibo. Subsequently, we construct a cross-platform user classification model, and realize cross-platform user classification and their behavioral patterns analysis, utilizing the K-means algorithm.[Findings/Conclusion] Cross-platform users can be categorized into three types: cross-platform bilateral heterogeneous users, cross-platform bilateral homogeneous users, and cross-platform unilateral active users. Among these, cross-platform bilateral heterogeneous users constitute the largest proportion, presenting different content on each platform based on their understanding of the platforms. Cross-platform bilateral homogeneous users show minimal variation in content presentation between the two platforms, but exhibit significant differences in interaction feedback. Cross-platform unilateral active users represent the smallest proportion, and are characterized by marked disparities in engagement levels and interaction values between platforms. [Originality/Value] This study develops a cross-platform user classification model and its corresponding indicator system, elucidates the characteristics of different types of cross-platform users and uncovers the across-platform bilateral development tendency of users. The research holds significant value in constructing user profiles and understanding their behavioral patterns in cross-platform contexts, and provides valuable insights for optimizing cross-platform ecosystems.

Keywords: Cross-platform, Knowledge exchange behavior, Cross-platform user classification, CRFM model, K-means