INSPECTING STUDENTS’ ACADEMIC REALITIES USING SOCIAL MEDIA PLATFORM

Priya Lande, Vipul Dalal

Abstract: The informal conversation of the students used on social media has been studied vastly now-a-days to understand their experiences. Twitter has been examined in which the hashtag #EnggProblems is mined to extract tweets and to classify the problems of the students into categories: Heavy study load, sleep problems, lack of social engagement, diversity issues, negative emotion and there is one more category that is ‘others’ .many classification algorithms were applied on the tweets extracted from the said hashtag, out of which naïve bayes multi label classifier gives the best results. It has been observed that most tweets get classified into others category as a result of which students’ problem cannot be identified clearly. This paper proposes an approach where SentiWordNet has been used to assign scores to each tweet and classify them as positive, negative or neutral. This helps to understand the sentiment of the student tweeting and identify at risk students

Keywords: Student, classification, SentiWordNet, education.

DOI: https://doi.org/10.15623/ijret.2016.0501045