Question Answering

This is one of the suggested topics for the course Language Technology Project 2005.

Introduction

A question answering system is a system which responds to to questions in natural language with short, precise and accurate answers. A lot of effort has been put in developing question answering systems for English which has resulted in systems like Start at MIT. Work on other languages is coordinated in the Cross Language Evaluation Forum.

Task

Develop and build a question answering system for a language other than English (preferably for Dutch). Because of the time constraints related to the project, the questions can be restricted to topic famous persons.

The suggested corpus for extracting information is the free online Wikipedia encyclopedia. The proposed implementation method for the question answering system is table lookup.

Modules

The question answering system will contain several modules which will be realized by different people in the project groups:

• Question interpretation module: reads a natural language question and returns relevant database queries (one or two persons).
  Relevant background knowledge: natural language processing plus basic programming skills.
• Retrieval module: collect documents from Wikipedia and extract relevant facts (pairs or triplets) from the web pages (two persons).
  Relevant background knowledge: retrieval and experience in programming of web interfaces.
• Database module: store facts (pairs or triplets) in a database, deal with incoming queries and return relevant answers (one or two persons).
  Relevant background knowledge: databases, SQL and basic programming skills.
• Answer formulation: convert the database answer to a answer in natural language (optional: zero or one persons).
  Relevant background knowledge: natural language processing plus basic programming skills.

The project members will need to start with deciding what type of questions they want the system to be able to deal with and what type of information to extract and store in the database.

Literature and tools

• PhD thesis Christof Monz (2003): chapter 1, chapter 2 and bibliography.
• Description of a traditional QA system by Julian Kupiec.
• Paper about the QA system Start at MIT.
• Tools and example scripts for common tasks