Date of Award

2015

Publication Type

Master Thesis

Degree Name

M.Sc.

Department

Computer Science

Keywords

Question Answering System

Supervisor

Morrissey, Joan

Rights

info:eu-repo/semantics/openAccess

Abstract

The search engine searches the information according to the key words and provides users with related links, which need users to review and find the direct information among a large number of webpages. To avoid this drawback and improve the search results from search engine, we implemented a Boolean based Question Answering System. This system used Boolean Retrieval Model to analyze and match the text information from corresponding webpages in the document indexing step when users ask a Boolean expression based question. To evaluate system and analyze Boolean Retrieval Model, we used the data set from TREC (Text Retrieval Conference) to finish our experiment. Different Boolean operators in the questions such as AND, OR has been evaluated separately which is clear to analyze the effectiveness for each of them. We also evaluate the overall performance for this system.

Share

COinS