A System for Extracting and Ranking Name Aliases in Emails

Mining potential information about person identity in emails is one of the popular research topics in email mining. This paper focuses on mining name aliases of a user from emails. Firstly, a system for extracting and ranking name aliases is proposed, which includes two modules: the Alias Extraction Module and the Alias Authority Ranking Module. Secondly, the methods used in the Alias Authority Ranking Module to rank the authority of name aliases of a user are presented in detail, which are based on email communication relation analysis and morphologically similar alias clustering. At last, we evaluate the proposed methods on the public subset of the Enron corpus. Experiment results show that the proposed system can efficiently extract name aliases and find the authoritative aliases of a user.