A sentiment analysis of blended learning experience and users' satisfaction across Twitter / Rozianiwati Yusof, Nor Asma Mohd Zin and Noorihan Abdul Rahman

• Main Authors: Yusof, Rozianiwati, Mohd Zin, Nor Asma, Abdul Rahman, Noorihan
• Format: Article
• Language: English
• Published: Universiti Teknologi MARA Cawangan Pulau Pinang 2017
• Subjects: Cooperative education; Computer assisted instruction. Programmed instruction; Intelligent tutoring systems
• Online Access: http://ir.uitm.edu.my/id/eprint/46600/1/46600.pdf; http://ir.uitm.edu.my/id/eprint/46600/

Social media has the power to spread information among users. It is very beneficial for improving education by giving valuable feedback or sharing perceptions on some related issues. Blended learning approaches are massively being used by most organisations, especially in higher education. However, there are many issues triggered among users when blended learning is implemented. Many researchers keep accomplishing new studies since this method has already been applied in many organisations and institutions. This paper aims to identify tweets regarding blended learning experiences. This study will use these tweets to define users' perception and acceptance of the blended learning method by producing a classifier model. This model will classify the tweets into positive and negative experiences in the blended learning approach. All the processes involved are implemented using Rapid Miner software. About 448 relevant reviews have been selected, and 84.38% of users give positive responses. The Naïve Bayes technique produced the best classifier model with 91.90% accuracy. The result defines most of the positive reviews are involving students, which are occurred 107 times, and 20 times for negative reviews. Meanwhile, the word “teacher” is appeared in 41 positive reviews and 14 negative reviews. This study suggests that blended learning should be continued because most users give positive reviews of the blended learning method. However, some acceptance criteria in blended learning should be considered such as age, gender, and region.