Efficiently denoising SMS text for FAQ retrieval
Ruhi Batra, Sanchit Sharma, A. K. Shrivastav, Puneet Kumar Goyal · 2014
Several online resources in the form of Frequently Asked Questions (FAQs) provide useful and much needed information across different domains like health, education, banking etc. Community based service has emerged as a powerful resource for information retrieval but its access is restricted to internet only.To access information without internet facility more and more people rely on Short Message Service (SMS) to get an instant answer to their query.Therefore, efforts have been put to improve the SMS based information retrieval system. The text in SMS messages are generally noisy and correcting this noisy text is one of the major challenges that affect the efficiency and accuracy of any SMS based information retrieval system. This paper provides the improvements to the existing algorithms of noise removal in SMS text to obtain better results. Experiments using different test cases show that the proposed system outperforms other methods.