Abstract:
This paper describes FAQ finder system to
search the user required questions. FAQ FINDER
is a natural language question-answering system
that uses files of frequently-asked questions as its
knowledge base. FAQ FINDER retrieves existing
ones found in frequently-asked question files. FAQ
FINDER uses a combination of statistical and
natural language techniques to match user’s
questions against known question answer pairs
from FAQ files. In this paper, this system will
implement one of the main parts of Information
Retrieval system: FAQ finder. It is also the
combination of statistical search and semantic
search. It uses Cosine similarity method to
calculate the similarity between user's question
and FAQ questions. Stemming: Porter Stemmer
and Eric Brill's Tagger are also used to process
the preprocessing steps. WordNet is also used to
take synonyms in order to perform as a semantic
search. Finally the performance measure is
calculated. For performance measure calculation,
system uses precision and recall methods.
