Efficient Retrieval of Frequently Asked Questions (FAQs)

Many methods and techniques have been developed to retrieve requiredinformation from bulk of documents. In this work we have comparedsome techniques for retrieving best matched FAQs (Frequently AskedQuestions) from FAQ list for given noisy SMS query. Noisy queries alwayscontain multiple noisy stop-words and less informative noisy words. It istrickier to get best correct word for a noisy word. Based on theassumption from which domain (like Railway, Health, Career etc.) a querytends to be, it becomes helpful to get absolutely correct words for all thenoisy words of a query. Probability theory becomes helpful in identifyingdomain of a query. We have applied those two techniques to get betterpossible words for noisy words and we got best results with ourexperiments.