WEIGHTING QUERY TERMS USING WORDNET ONTOLOGY

Term-weighting techniques are responsible for giving a weight for each query term to determine its significance. This term significance is based on the generality or specificity of that term. Traditional weighting models, such as the Inverse Document Frequency (IDF), measure the generality/specificity of terms through document collection statistics. This paper presents a technique that employs the WordNet ontology to determine the significance of query terms without depending on document collection statistics. The experiments are carried out on the WT2g document collection under the Terrier Information Retrieval Platform.