Word Sense Disambiguation for Semantic Applications

Natural language processing (NLP) has become the most significant obstacle that has been restricting the applications via the web. Today, very little of the content on the web can be understood by the machines, although vast amount of electronic information has been kept on them. Word sense disambiguation (WSD) is an important intermediate step in many language processing applications. It is basically a mapping function from the context to the set of senses. This function has many parameters that are difficult to explore. The factors effecting the success of WSD systems are generally very sensitive to these parameters. The issues in WSD are examined in the context of Turkish WSD application. Ambiguous words and their sense classifications have been established and by providing manually sense tagged corpora and examining WSD problem from various perspectives, an important contribution has been achieved for the researches in this domain.