Abstract:

The talk will give an overview of the approach taken at the FernUniversitaet in Hagen
to develop systems for the automatic semantic analysis of natural language.
After a short introduction to the underlying semantic framework MultiNet, I
will focus on the structure and development of the computational lexicon HaGenLex,
a general domain lexicon for German that contains morpho-syntactic and semantic
information including semantic role frames and selectional restrictions. HaGenLex
serves as the central lexical resource for deep semantic analysis, i.e., for
transforming natural language expressions into MultiNet representations. In the
final part of my talk, I will turn to applications of the semantic analyzer to
Question Answering tasks.