Recent works in computational linguistics have investigated the association of a text with a certain domain(e.g. Sport, Medicine, Politics, …). Using such associations has been shown a significant improvement in the performance in tasks such as word sense disambiguation and lexical acquisition. We propose an unsupervised methodology for the estimation of the relevance of a domain in a text. The method combines the knowledge in WordNet Domains, an extension of WordNet in which synsets are annotated with domain labels, and a probabilistic framework which makes use of a balanced corpus to induce domain frequency distributions
Unsupervised Domain Relevance Estimation for Word Sense Disambiguation
Gliozzo, Alfio Massimiliano;Magnini, Bernardo;Strapparava, Carlo
2004-01-01
Abstract
Recent works in computational linguistics have investigated the association of a text with a certain domain(e.g. Sport, Medicine, Politics, …). Using such associations has been shown a significant improvement in the performance in tasks such as word sense disambiguation and lexical acquisition. We propose an unsupervised methodology for the estimation of the relevance of a domain in a text. The method combines the knowledge in WordNet Domains, an extension of WordNet in which synsets are annotated with domain labels, and a probabilistic framework which makes use of a balanced corpus to induce domain frequency distributionsFile in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
