Forecasting the behaviour of Trending Terms in Microblogs

Abstract

Due to the tremendous usage of social media platform, its need of an hour to understand the societal opinion trends about certain topics posted on social media. Monitoring and understanding the pulse of social media allows to gain feedback and to understand the people or market in a better way. There is also an outpour of information, related to the thoughts, feelings, expressions, experience, likes and dislikes that everyone tends to give out in social media. The catch is to utilize this for a novel purpose that can aid the thinkers, technocrats, leaders, communications, innovators to build a proactive say today, on the challenges of tomorrow. It will be like learning the ‘mind of the society’ and to aid the citizens with the necessary conditions to make their lives better. This is exactly what we have attempted to do here.By digging deep into the data generated by social media, we can get the answer of following questions; What happened in the past? 2) What is happening now? & based on these answers we can obtain the answer to the question What can happen in the future?. We have put forth a more accurate and refined prediction algorithm which deals with the weight of the term and instrument Poisson distribution function to forecast behaviour of the term for next cycles.