Facts or Feelings - A Novel Framework For Classifying Survey Text Data
Debanjan Banerjee · International Journal of Modern Trends in Engineering and Research · 2015
The rapid growth of online survey responses on important social issues gives birth to the need for automated tools for analysing potentially beneficial patterns and trends for the society at large. In this paper we propose a framework to classify online survey responses as either intelligent or egoistic. We classify those survey responses as intelligent if they are based upon objective facts. Survey responses are classified as egoistic if these are based upon subjective considerations. We believe this approach will pave the way for future works to create a substantial approach for extracting knowledge out of textual data. Keywords - Text classification, Survey text classification, Intelligent and egoistic classifications, Machine learning I. INTRODUCTION The humongous growth in Internet communications has given rise to the phenomenon of online surveys. Increasingly people opine about their preferences and attitudes on various important social events such as sporting occasions and elections through online surveys. The sporting events such as the soccer world cup generate a lot of interesting views amongst these survey responders. Many marketing agencies can boost their marketing policy making once they are able to tap into these comments as this would be helpful to understand about popular mood over a well-known event. The primary challenge for entrepreneurs and researchers alike while dealing with survey reviews is to bring about organized frameworks that turn untidy and unstructured survey responses into meaningful information. Once the researcher community is able to come up with these frameworks it will enrich the efforts to understand how humans influence each other during popular social events through survey responses. The present work intends to aid this endeavour by classifying these survey responses into two distinct classes which are intelligent and egoistic. The survey questions have been designed with the objective of gauging popular opinion on the recently concluded World cup soccer in Brazil on June-July, 2014. This work will present a framework which will look into classifying the survey responses into two distinguishable classes i.e. intelligent and egoistic. This paper will first document in detail the past work which has been performed in the same area and will provide arguments to distinguish it from these past works. Then the work will detail the working process involved alongside the mathematical formulations which have been considered on the work. The results and the conclusions which follow from the results will also be documented later on.