Find spammers by using graph structure

摘要

In recent years, many social network services have made lots of business models. However, it has big weakness related with personal information spill or attacks from spammers. There are many conventional approaches about spammer detection. Almost the whole studies suggest solutions by using vertex degrees or Local Clustering Coefficient. However, they have been caused false positive results. In other words, many normal vertices can be mistaken for spammers. Therefore, we propose a new approach by using the circuit of graph structure. And then we demonstrate the strengths of our work by employing the experiment results.