Supervised Learning Techniques for Classification Of Students’ Tweets

Main Article Content

Blessa Binolin Pepsi M, et. al.


In today’s era, up-to-date information can be retrieved from social network, internet community and data forums. People especially the younger generation share their feelings, happiness, experience and also day to day happenings in the social media platforms like Twitter. There exists large volume of unstructured data in it. The proposed system concentrates on the learning process of the engineering students and the problems faced by them during their study from their twitter posts. Since the data collected is huge, Apache hadoop map reduce environment is used for processing. The system includes pre-processing of tweets, calculating F1 measure, identifying prominent categories, identifying word and category probability and finally classifies tweets to the respective categories. The supervised learning techniques such as multiclass SVM based Platt Scaling, Naïve Bayes and logistic regression are used to identify heavy study load, lack of social engagement and sleep problems. Comparing the results attained, SVM achieves an accuracy score of 84% which is 5 to 10 percent higher than Logistic Regression and Naïve Bayesian method.


Download data is not yet available.


Metrics Loading ...

Article Details

How to Cite
et. al., B. B. P. M. . (2021). Supervised Learning Techniques for Classification Of Students’ Tweets. Turkish Journal of Computer and Mathematics Education (TURCOMAT), 12(12), 3110–3118.