Extracting semantic descriptions of expertise from researcher publications

Proximify is a web services company which provides software that supports the creation of privatecollaboration networks. Connections are identified among members of the network on the basis of sharedinterests, skills and experience. It also supports different views of the organization - including maps,publications and intelligence clusters - combined with filters to customize the views to the user's requirements.One of the challenges is the extraction and representation of members' expertise. The current version of thesoftware requires each member to manually select areas of expertise, or define her own. This can lead tosuboptimal characterization of expertise, with multiple wordings of the same expertise, making difficult theautomatic identification of connections. The proposed project will investigate novel representations ofexpertise that are extracted automatically from the publications or reports a member has authored. Theserepresentations are semantic, in that they capture the meaning of the publications and are robust in the presenceof vocabulary mismatch between member expertise profiles. The semantic representations will support theimportant functionality of matchmaking between industry and faculty members, based on a free text descriptionof the company's interests, and interactive visualizations of expertise landscapes of units and institutions.