Natural Language Processing Research

Abstract

Question-Answering In this talk, I will give an overview of my current research in Question-Answering (QA) systems. QA systems are considered as one of the major breakthroughs in Computer Science. In the recent years, they have received significant attention from academia and industry. For example, the IBM Watson QA system won the prize money of the Jeopardy show by outperforming human champions in answering quiz questions. Despite considerable progress, several challenges still exist in the design and implementation of QA systems. Most notably, the knowledge acquisition phase of classical QA systems relies on deep Natural Language Processing (NLP), such as syntactic parsing and co-reference resolution, to infer facts from huge volumes of texts. However, these sophisticated NLP techniques are not always accurate, leading to error propagation in the pipelined QA architecture, and ultimately compromising the overall performance. Furthermore, deep NLP is computationally expensive, which hampers the applications of QA systems in real-life, practical applications.

Cite this paper

@inproceedings{Ittoo2014NaturalLP, title={Natural Language Processing Research}, author={Ashwin Ittoo}, year={2014} }