Analysis of Public Sentiment Variations on Twitter Data Set

Abstract

Abstract

Large numbers of users share their opinions on social networking sites, making it a valuable platform for tracking and analyzing public sentiment. Twitter has become a valuable platform for public to share their opinion. It can be used for sentiment analysis. Such tracking and analysis can provide critical information for decision making in various domains. It has attracted attention in both academia and industry. Sentiment analysis is nothing but finding about what any kind of text wants to convey. It is nothing but opinion mining. Previous research mainly focused on modeling and tracking public sentiment. But in this, moved one step further to analyze public sentiments. Here observed that emerging topics (named foreground topics) within the sentiment variation periods are highly related to the genuine reasons behind the variations. Based on this observation, proposing a latent Dirichlet allocation (LDA) based model, foreground and background LDA (FB-LDA), to distil foreground topics and filter-out longstanding background topics. These foreground topics can give potential interpretations of the sentiment variations. Parallel processing is applied to propose model which will reduce the time required for processing the number of tweets and ultimately improve the performance. 

Keywords: Twitter, public sentiment, emerging topic mining, sentiment analysis, latent Dirichlet allocation, Gibbs sampling

Cite this Article

Ankita Bondre, S.V. Hemanth. Analysis of Public Sentiment Variations on Twitter Data Set. Recent Trends in Parallel Computing. 2017; 4(2): 33–58p.