Research on User Clustering Algorithm Based on Software System User Behavior Trajectory
Yanping Tang, Bingxian Ma, Zheng Wu · 2019
This paper studies and analyzes software system user behavior, proposes a clustering algorithm based on user behavior trajectory data. Through a series of data processing steps, the original user access and operation data are transformed into user behavior trajectory matrix, and the behavior trajectory matrix is normalized. Based on the generated user behavior trajectory matrix, the user behavior information contained in the matrix is quantified, and the concrete steps of user behavior clustering algorithm are proposed. According to this clustering algorithm, the similarity between user behavior trajectories can be calculated, and the users can be clustered according to the set similarity threshold. For users clustered in the same type, these users have similar access or operation habits. Experiment shows that the algorithm has good clustering effect and applicability.