Information Retrieval by Modified Term Weighting Method Using Random Walk Model with Query Term Position Ranking

Term weighting is a core idea behind any information retrieval technique which has crucial importance in document ranking. In graph based ranking algorithm, terms within a document are represented as a graph of that document. Term weights for information retrieval are estimated using termpsilas co-occurrence as a measure of term dependency between them. The weight of vertex in the document graph is calculated based on both local and global information of that vertex. This paper introduces a method of information retrieval using random walk model considering positional values of a term in the document for computing its inverse document frequency and assigning trained weight to terms in the user provided query. Experiments on standard datasets have shown that our approach provides improvement in recall and precision of information retrieval system.