Document Retrieval Based on Question Answering System

This paper presents our methods in research for building a search system model based on natural language interaction, specially used in e-library. Main aim of this system is to support users to enter simple English queries for searching information about the books, such as title, author, language, category, publisher... In this system, we focus on solving majority problems to process the natural language query: approaches of syntax analysis and syntax model, semantic model, transformation mechanism from semantic model into database queries... We built a specific search system for the free e-library Gutenberg. The results of our experience are effective and suitable to develop the other search systems which work on descriptive information about documents, courses, books in e-libraries.