Enhancing the relevance of information retrieval by querying the database in natural form

Abstract

Effectiveness of Information Retrieval from Structured Database through Natural Language provides high utility value and ease of use. The challenge in Information Retrieval is to derive the users intended from limited number of query words (sometimes just one word) to extract the relevant information from structured database. It calls for correct interpretation, disambiguation and context resolution in natural language query processing. These difficulties can be reduced by limiting the domain of applications. In this paper we present our work in designing and developing Student Interface System by querying user in Natural Language (English). Our model consists of (i) semantically parsed the user inputted sentence (ii) transforming it into intermediate query form and finally (iii) convert it into structured query language. The system is developed in JAVA using JDBC for Student Information System. The same approach and methodology can be used in developing queries using vernacular languages.