COMPREHEND PUPIL ISSUES BY SOCIAL MEDIA MINING

Abstract

Young people spend most of their time on the social media sites such as Twitter, YouTube, Facebook etc. Most of the day, they spend on these sites texting, posting, tweeting about their inner thoughts. There is an abundant amount of information on social media that is untapped. Hence Social media mining is an efficient way to understand the emotions of our young students go through.  Using twitter, we can extract tweets according to geolocation and analyse the data to deduce these social issues of students facing at that particular location. The data we get from social media is dirty. We have to clean that data and calculate the sentiment-score by using the best algorithm. Using Sentimental score, we observe whether a tweet reflects a positive, negative or neutral sentiment. In our study we have created a deep learning model which figure out the sentiment of the tweet. Using retweet count and favorite count as parameter we deduce the most negative tweets, thus helping us to address the most relevant issues faced by the students in educational institution.