RESULT PREDICTION FOR POLITICAL PARTIES USING TWITTER SENTIMENT ANALYSIS

Elections in general is a decision making process by which people choose an individual to hold power of public office. Recently, social media has a huge impact on people's opinion towards political parties, candidates and elections. Twitter is a blogging service which has millions of tweets over particular topic. Moreover, each tweet may have either a positive, negative or neutral sentiment towards the topic, which needs to be considered. We use different machine learning algorithms like Naive Baye's, Support Vector Machine both linear and radial basis kernel function. By gathering a large dataset of more than 500,000 tweets out of which 80% of data is taken for training algorithms and remaining 20% for testing. By applying this technique to Indian parliamentary and state elections we predict which political party has more influence on social media