This document describes the work performed by the TALP Research Center, UPC in its first participation at ERD 2014 short text evaluation track. The objective of this evaluation track is to recognize mentions of entities in a given short text, disambiguate them and map them to the entities in a given collection of knowledge base. To this end, we presented our system taking advantage of a topic modeling approach to rank candidates of each entity mentions occurring in the query text

Peer Reviewed