Abstract:Online communities have developed into a gathering place for knowledge demanders to continuously interact and derive knowledge. It is important to explore the behavioral patterns of knowledge demanders' participation in knowledge mobilization in online communities in order to play and promote knowledge value maximization. Firstly, the time series method is used to analyze the distribution pattern of user activity in the temporal dimension; secondly, the social network method is used to explore the participation and mobilization relationship of knowledge demanders in the spatial dimension; finally, the rooting theory coding technique is used to analyze the interactive content and construct a binary structure model of knowledge mobilization in online communities in the content dimension. The results found that: knowledge mobilization activities in online communities are driven by participants and have member spontaneity and self-adaptation; knowledge mobilization in dual structure has obvious knowledge asymmetry, and knowledge producers and knowledge demanders are unevenly distributed, resulting in knowledge discrepancies; successful operation of online communities needs to rely on enhanced knowledge mobilization by core participants of activities and de-marginalization of passive participants, so as to drive the knowledge mobilization ability of the whole online community users and form knowledge empowerment and promotion to strengthen the knowledge mobilization effect.