A Proposed System for Sentiments Prediction through Social Media Data Analytics

Nowadays people are very active and open about expressing themselves on social media due to the liberty provided there. People express all kinds of emotions honestly. Sentiment Analysis is the process of identifying and categorizing opinions and views expressed by humans in a piece of data. This is useful to gain an overview of people and their thought process. This analysis can be put to use in many areas like entertainment, sports, health care, etc. In our implemented system we extract tweets using Twitter API and python functions. We create a dataset from these tweets and give them to Knime tool as training and test dataset. Using different Sentiment Analysis algorithms like Naive Bayes, Decision Tree, etc. the Knime tool gives us output as classified data. The data is classified into 2 polarities: Positive and Negative. The tweets containing words such as happy, excited, great, etc. are assigned with a positive polarity while those having words like sad, sick, worried, etc. are assigned with a negative polarity. Using these polarities, we can predict the moods of the sports players and help improve them so that it doesn't affect the performance. We can use it for movie reviews and also for feedbacks for services of hospitals.