Abstract
Question-answering (QA) research emerged as an attempt to tackle the information-overload
problem. Question answering systems take users natural language questions and locate answers from
large collections of documents. The current existing Question-Answering systems can deal with shallow
questions (factual questions) somehow easily and correctly, but asking deep questions (non-factual
questions or complex questions) such as Why-questions are more difficult than asking shallow questions.
However, these complex questions cannot be neglected as input for a QA system as they comprise about 5
percentage of all why-questions and it shows that this kind of questions do have relevance in QA
applications.
The techniques that have proven to be successful in QA for factual or factoid questions that
returns answer as a noun phrase are found to be not suitable for questions that expect an explanatory
answer. The work made an analysis on different question answering methods and found that discourse
structure based method is effective in dealing with non - factual questions. So the work discuss about the
role of dscourse structure in dealing with ’why’ questions since it is based on Rhetoric Structure Theory,
which helps in identifying the relationship between sentences or paragraphs from a given text or
document and investigate to what extent discourse structure does indeed enable answering whyquestions
Users
Please
log in to take part in the discussion (add own reviews or comments).