Dealing with word sense disambiguation in lexical chaining

Speaker: Mattt Enss

A lexical chain is a sequence of words in a document that are semantically related (i.e., related in meaning). Lexical chains indicate where certain topics or subjects are being discussed in a document. The chains therefore can provide context and be used to determine where topic changes occur. Other applications include document summarization and keyword generation for documents.

Prior work of mine focused on providing a generalized framework for lexical chaining algorithms. This framework divides lexical chaining into two distinct problems: performing word sense disambiguation, and determining the semantic relations between disambiguated words. Adapting prior lexical chaining algorithms, all of which perform word sense disambiguation while chains are being built, to this framework reduces the number chaining mistakes made. My current work focuses on the results observed when a prior lexical chaining algorithm (Silber and McCoy 2002) is adapted to my framework, as well as tackling some of the problems involved in using WordNet to determine semantic relatedness.

My talk will start with an overview of lexical chaining, word sense disambiguation and semantic relatedness, before moving on to my previous and current work.