Two Web-Based Approaches for Noun Sense Disambiguation

The problem of the resolution of the lexical ambiguity, which is commonly referred as Word Sense Disambiguation (WSD), seems to be stuck because of the knowledge acquisition bottleneck. It is worthwhile to investigate the possibility of using the Web as a lexical resource. This paper presents two attempts to use the Web not for extracting training samples but for helping during the WSD task. These two approaches investigate the effectiveness of using the redundancy of the Web to disambiguate nouns in context as well as using modifier adjectives as supporting evidence. Preliminary results show that the direct use of Web statistics allows only for the adjective-noun pairs approach to obtain a better precision than the baseline, even if with a low recall. The evaluation was carried out with different search engines, and the obtained results were almost identical. Finally, the Web was more effective than the WordNet Domains when integrated rather than stand-alone.

• Type: Conference
• Year: 2005
• Where: CICLING
• Authors: Paolo Rosso, Manuel Montes-y-Gómez, Davide Buscaldi, Aarón Pancardo-Rodríguez, Luis Villaseñor Pineda
• Keywords: CICLING 2005, Knowledge Acquisition Bottleneck, Lexical Ambiguity, Natural Language Processing, Word Sense Disambiguation