Word Sense Disambiguation and Induction
Posted under: Course and Workshop Descriptions; Language and Computation.
Word Sense Disambiguation (WSD), the ability to identify the intended meanings of words (senses) in context, is a key problem in Natural Language Processing (NLP), potentially enabling deeper representations of text. WSD is performed with respect to an existing inventory of word senses. If such inventory is not available, or within application-driven scenarios, a sense inventory can be automatically acquired from text corpora, a task known as Word Sense Induction (WSI).
This course will provide an introduction to Word Sense Disambiguation and Induction. The aim of the course is two-fold: first, we introduce the audience to a wide range of techniques for the two tasks; in addition, we provide tools for the development of systems able to participate in past and current evaluation exercises for WSD and WSI (Semeval-2007 and 2010).
The course will be structured as follows.
1) Introduction: WSD and WSI
2) WSD: methods and issues
3) WSI: methods and issues
4) Evaluation measures and campaigns (Semeval-2007 and 2010)
5) Lab: in-depth analysis of a sample system
Part of the course will follow and extend material appearing in a recent survey article:
R. Navigli. Word Sense Disambiguation: a Survey, ACM Computing Surveys, 41(2), ACM Press, 2009, pp. 1-69.
http://www.dsi.uniroma1.it/~navigli/pubs/ACM_Survey_2009_Navigli.pdf
Further particulars (e.g, prerequisites):
This tutorial targets both computer scientists and computational linguists. The tutorial is self-contained, so no specific background knowledge is required. The objective is to give all attendees a clear understanding of a wide range of techniques for WSD and WSI.
