Paper Title
Context Based Document Retrieval Using Taxonomic Tree Structure

Document retrieval is task of retrieving relevant documents according to user query. This paper solves a major problem in text based information retrieval, using the limited set of query terms given by user and extracting small set of relevant documents in a large document set. We propose search queries formulation for each query entity based on some supplementary information so that to detect target entity from the set of retrieved documents. Proposed method utilizes contextual information formed by word relationships contained in the document collection, and auxiliary information is given by context terms evaluated through the expression of a relation. To facilitate the specification of search queries and document comparison at retrieval time, user key words are compared against taxonomic tree structure computed using context database. If the retrieved document is less than top K documents KNN algorithm is applied on taxonomic tree structure, to retrieve top K relevant documents. Document relevance is determined by comparing the similarity of search and document tree structure. The disclosed system attains greater search precision by retrieving relevant documents based on context and facilitating easier search specification. Keywords—Information retrieval; context term;term frequency; Taxonomic tree structure;Relation extraction; Relation completion