Determine Word Relevance in Document Queries Using TF-IDF
Ashok Koujalagi · International journal of scientific research · 2015
In this paper, we examine the results ofapplying Term Frequency Inverse DocumentFrequency (TF-IDF) to determine what words in a corpusof documents might be more favorable to use ina query. As the term implies, TF-IDFcalculates values for each word in a document throughan inverse proportion of the frequency of theword in a particular document to the percentageof documents the word appears in. Words withhigh TF-IDF numbers imply a strong relationship with the document they appearin, suggesting that if that word were to appear ina query, the document could be of interest tothe user. We provide evidence that thissimplealgorithm efficiently categorizes relevantwords that can enhance queryretrieval.