Knowledge creation opportunities for the K to 12 educationaltransformation in the Philippines using predictive analytics

Abstract

This research runs a predictive model using a simple decision tree of the twitter mined about the opinion of the people towards the new K-12 education program of the Philippines. The initial study which acquired sentiments from Twitter microblogs was utilized to find out whether a predictive model can substantially generate knowledge to support the K to 12 educational reforms in the Philippines. RapidMiner was used as tool to perform analytics on Twitter data. Various RapidMiner operators were used to process the Twitter microblogs to perform clustering and predictive analytics. It also utilized AYLIEN as an extension module of RapidMiner for text analysis and extract insights from these tweets. The experiment reveals in word cluster analysis that users who expressed sentiments about K-12 used similar words on the messages they posted. Overall, the results suggest that tweet data have a quite peculiar nature. Words used in discussed topic create a sort of Twitter culture. The results showed that in the decision tree generated, only favorites variable or the number of likes on a K-12 tweet provides a strong indication of classifying a K-12 tweet as subjective or objective.