Twitter truths: Authenticating analysis of information credibility

With the growth of internet, life of people have also changed. People share a huge amount of information on social media, thus increasing the amount of information at a very fast rate. This has made Social Data Analytics an important area of research. Analyzing the opinions or sentiments of people on the data shared can serve as a great means for making decisions in various domains. Also, not each and every data present on social media is correct as various people try to spread rumors, mostly just for fun. Due to this, it is required that there exist methods using which people can easily check information credibility of data concerning social media. The research carried out till now have mainly focused on tracking and modeling sentiments of public. This paper proposes a method for assessing the credibility of information by emulating the methods of topic and opinion classifications, primarily focusing on Twitter data. Improved Latent Dirichlet Allocation (LDA) algorithm has been used for classification of topics within the tweets and semi-supervised method (SVM) has been used for sentiment analysis. Finally, majority decision by comparing the number of contrastive opinions about a topic is performed, for credibility analysis. Though, English and Hinglish languages have been used here, however, this method can be adapted easily to other languages since Twitter API allows specifying the language of the retrieved posts.