Discovering and Disambiguating Relevant Entities in Text

The Faculty of Informatics is pleased to announce a seminar given by Salvatore Trani

DATE: Wednesday May 13th, 2015
PLACE: University of Lugano, room SI-015, Informatics building (Via G. Buffi 13)
TIME: 13.30

ABSTRACT:
Discovering and disambiguating relevant entities mentioned in a natural language text force the resolution of ambiguous names to concrete entities registered in a given knowledge base such as Wikipedia, DBpedia, or Freebase. Knowing the specific entity is an important asset for several other tasks, e.g. entity-based information retrieval or higher-level information extraction.

I will briefly introduce you to the Entity Linking problem, describing also the most popular solutions proposed in literature. Then, I will present our contributions to the field: (i) a new relatedness measure which is one of the core features used in disambiguation, (ii) a modular open source framework for entity linking, (iii) to go beyond entity linking by trying to accurately estimates the most relevant entities mentioned in a document (a.k.a. entity saliency).

BIO:
Salvatore Trani is a PhD student at the Computer Science department of the University of Pisa. He is also collaborating with the Italian National Research Council (CNR) of Pisa. His current research focus is the linking of unstructured text to structured knowledge bases by discovering and disambiguating entities, as well as the use of entities and knowledge bases in information retrieval tasks.

HOST: Prof. Fabio Crestani