Detecting Sentiment Fluctuations in Multimedia Learning Material


ÖZGÜR A., Saglam F., GENÇ B., ALTUN A.

PAMUKKALE UNIVERSITESI EGITIM FAKULTESI DERGISI-PAMUKKALE UNIVERSITY JOURNAL OF EDUCATION, sa.60, 2024 (ESCI) identifier identifier

Özet

With the emotional design of multimedia learning materials (MLM), goals such as creating a certain emotion in learners (positive-negative), regulating learners' motivation, influencing their cognitive characteristics, and learning outcomes are pursued. Thus, it is important to develop certain guidelines to ensure the affective quality of the MLMs. In this study, the Sentiment Map Model (SMM) was used to detect the sentiment fluctuations of two MLMs with positive and negative emotional designs in terms of their texts. SMM is a lexicon-based sentiment analysis tool. As the first step in the SMM process, the sentences of positive and negative MLMs were determined. Then, using SWNetTR++ lexicon, the sentiment tones of the sentences were calculated. The calculated sentiment tone values for positive and negative MLMs were placed on the Shewhart Control Diagram and sentiment fluctuations of the sentences were visualized. Four analysis rules (stable, significant, strong and violent) were applied to find consistency regions in the sentiment fluctuations and those regions were highlighted in the diagram. As a result, it was observed that there were more stable sentiment fluctuation regions in the positive MLM than in the negative MLM. In this context, the sentiment analysis of the texts in the MLMs with SMM, the emotional design of the MLMs and their use in learning -teaching processes were discussed.