A Community Detection Method Based on User Behavior

Abstract

The invention proposes a community detection method based on user behavior, which improves the existing method. The detection method includes: analyzing the user's historical playback records to obtain similar playback behaviors; mapping each record to a two-dimensional plane to complete the mapping of data to cells; aggregating records with similar playback behaviors into data piles, and Extract a part of representative playback behaviors as its playback mode; match the user's current playback behavior, determine which playback mode the behavior belongs to, and combine nodes with similar playback behaviors (that is, belonging to the same playback mode) to form Community. Through the implementation of this scheme, the records in each data pile have similar playback behaviors, which is more conducive to collaborative acquisition of video content among users and efficient sharing of local resources. In addition, clustering based on playback modes makes the feature description of user behavior more accurate. Accurate and high stability.

Description

A Community Detection Method Based on User Behavior technical field The invention relates to the communication field, in particular to a community detection method based on user behavior. Background technique With the development of mobile wireless communication technology, various real-time, dynamic and heterogeneous streaming media services have gradually become an important force to promote the development of the future Internet. Especially in Vehicle Ad Hoc Networks (VANETs), the emergence of new technologies has promoted the gradual increase of various streaming media applications based on VANETs, ​​and the user experience of Video on Demand (VoD) has also gradually increased. Therefore, service providers urgently need a video-on-demand service that supports interaction, collaboration, and sharing to speed up content distribution efficiency, ensure reliable service quality, and improve the utilization of network resources while meeting various needs of users. In ord...

AI Technical Summary

Problems solved by technology

However, since this method is a hierarchical clustering method, it has two deficiencies: on the one hand, it cannot accurately evaluate the interest similarity of nodes, and on the other hand, it cannot clearly define the boundaries of communities. Seriously lead to a decline in content retrieval efficiency, not suitable for current VoD services

However, the former has two deficiencies: on the one hand, due to the differences between sample nodes, the data and computing resources cannot be evenly distributed well;

However, hierarchical clustering relies on an evaluation method of node similarity, which cannot clearly define the boundaries of communities through the interaction between nodes in a P2P network environment, resulting in a decrease in the accuracy of aggregation.

Images