Personal Name Matching: New Test Collections and a Social Network based Approach

This paper gives an overview of Personal Name Matching. Personal name matching is of great importance for all applications that deal with personal names. The problem with personal names is that they are not unique and sometimes even for one name many variations exist. This leads to the fact that databases on the one hand may have several entries for one and the same person and on the other hand have one entry for many different persons. For the evaluation of Personal Name Matching algorithms test collections are of great importance. Therefore existing test collections are outlined and three new test collections, based on real world bibliographic data, presented. Additionally state-of-the art techniques as well as a new approach based on semantics are described.