Mining and Classifying Social Network Data: The Case on King Abdul-Aziz University Twitter Accounts

Abstract

Social media, and especially Twitter, in specific domains such as healthcare, education and politics, turned into the key venue of social interaction today. Many higher education institutions (HEIs) seek therefore to utilize the value-added of Twitter to disseminate and to collect information from students in view of improving the quality of education at their institutions. In this view, the ability to mine, classify and interpret the content of Tweets is crucial. By examining the case of the King AbdulAziz University in Saudi Arabia, this paper offers a preliminary insight into what kind of information can be collected and in which ways it can be useful for a given HEI as regards teaching, administration and overall management. To this end, this paper examines the usability of three machine learning models, including Support Vector Machine (SVM), K Nearest Neighbor and, finally, Random Forests (RF). The outcomes of this study that this paper elaborates on suggest that in terms of accuracy, SVM is the best performing classifier. Meanwhile, even if the RF proved to be a strong classifier too, it did not perform as well as the SVM.