A Test Collection for Email Entity Linking

Most prior work on entity linking has focused on linking name mentions found in third-person communication (e.g., news) to broad-coverage knowledge bases (e.g., Wikipedia). A restricted form of domain-specific entity linking has, however, been tried with email, linking mentions of people to specific email addresses. This paper introduces a new test collection for the task of linking mentions of people, organizations, and locations to Wikipedia. Annotation of 200 randomly selected entities of each type from the Enron email collection indicates that domainspecific knowledge bases are indeed required to get good coverage of people and organizations, but that Wikipedia provides good (93%) coverage for the named mentions of locations in the Enron collection. Furthermore, experiments with an existing entity linking system indicate that the absence of a suitable referent in Wikipedia can easily be recognized by automated systems, with NIL precision (i.e., correct detection of the absence of a suitable referent) above 90% for all three entity types.