Consolidating Identities of Authors through Egonet Structure

Individuals often appear with multiple names when considering large data-sets collected from different sources, giving rise to the name ambiguities. Name ambiguity comes in two flavors: a single individual appearing with more than one name (synonym problem); a single name being used to refer to more than one individual (homonym problem). This works focuses on the synonym problem and explores structural patterns of the collaboration network. We consider a scenario where the individual has a specific profile page that lists its bibliographic records. Despite more restrictive, this scenario has become quite common among digital libraries such as DBLP1 and Google Scholar2. Such profiles often have name ambiguities as the profile owner appears with different names within the bibliographic records (e.g., different name spellings, typos, etc). This paper tackle these name ambiguities without resorting to any content information!