A fuzzy inference supportive social media market analysis for predicting crowd influence in national elections

The interest of the crowd plays a vital role at national level elections not only to predict the influence of participating parties but also to make marketing strategies for the campaign. Beyond person-to-person interaction, social media is also used by society to put their views, comments, and interest toward political parties and their candidates. But, due to the large volume, variety, and velocity of social media data, it becomes difficult to analyze. Moreover, the extracted information from social tweets also raises issues of biased information. Here, in this work, a fuzzy inference supportive framework is proposed to study crowds’ preference on social media platforms to make campaign strategies by the political parties in national-level elections. The proposed approach utilizes tweets from Twitter, LinkedIn, and Instagram to predict the crowd’s influence on political parties. Further, the approach can be utilized by the parties to change the campaign strategy. In experimentation, the approach is tested on the real-time dataset collected from the social network websites, and found that the proposed approach is performing well to predict the user’s interest. Also, when compared with existing methods, the proposed approach’s performance is found significant over other methods.