Mining social media data of Philippine higher education institutions using naïve bayes classifier algorithm

Abstract

Higher Education Institutions in the Philippines integrate social media like Facebook to market their brand, give some announcements, update news and other important information about their institution. However, their stakeholders create unofficial Facebook pages associated to their institution. These pages are areas for students mostly to react to posts, comment, give their opinion on a given topic and share their daily life online in an informal and casual way. These social media footprints were used in the data mining process to identify their discourse. With this, Universities and Colleges could form decision-making mediations and improve the quality of education and service they provide. The specific objectives of this study are: to create a workflow on how to analyze social media data; to use Naïve Bayes Algorithm in classifying the discourse of Philippine HEI stakeholders; and to deduce the results of data mining and recommend intervention activities to improve quality education. Knowledge Discovery Databases (KDD) was used in the study. The result of the data mining process is displayed through a word cloud and a dynamic graph visualizing the classifications of posts and comments. Social engagement, academics, emotions, health, policies and finances were identified as the SM discourse of Philippine HEI Stakeholders.