Computational Linguistics Colloquium

Tuesday, 11 June, 14:15, Dekanatssitzungssaal, Building 10
NOTE UNUSUAL DAY, TIME and PLACE

The FrameNet project and its applications

This talk gives an overview of the FrameNet project which is housed at the International Computer Science Institute at UC Berkeley. FrameNet is in the process of developing a corpus-based lexicon of several thousand English lexical units described in terms of Frame Semantics (Fillmore, 1982).More specifically, FrameNet creates, based on word uses in large corpora, a database of lexical entries for English verbs, nouns, and adjectives taken from a variety of semantic domains. Each lexical entry in the FrameNet database provides an exhaustive acount of the syntactic and semantic combinatorial properties of each lexical unit (i.e., one word in one of its uses).

The talk is structured as follows. Part one outlines the goals of the FrameNet project. Part two summarizes the main principles of Frame Semantics, which underlies the analysis of words in FrameNet. Part three of the talk discusses the work flow of the FrameNet project (Frame composition, corpus extraction, manual semantic annotation, creation of lexical entries). Part four discusses how words from the commercial transaction and crime domain are analyed by FrameNet. Part five shows how information provided by FrameNet is useful for applications such as information retrieval, question answering systems, and machine translation.

If you would like to meet with the speaker, please contact Katrin Erk.