abstract
- Emotions classification or valence extraction in textual datasets, e.g.The students' opinion data, is becoming an emerging topic aimed at understanding the impact or intensities of words (terms) used by the users in the different contexts/applications. The method (emotional valence) for textual data scrutiny and exploration have proved to be central to the human experience analysis. To this effect, this study implements a text mining approach that determines the impact of the emotional valence (textual data quantification) shown by the students in their feedback provided during the semester course (pre and post) to determine its (emotional classification) relation or interconnectedness with the students' learning outcome and performance. The proposed method is designed based on the appraisal theories and component process model (CPM) that studies the degree of pleasantness or goal achievement as an effect of valence judgements. Technically, the study identifies the top terms in the students' data that can be used to draw insights or understand the learning experiences or performance by using Corpus feature selection and Term document matrix word processing libraries in R programming software. Also, it utilized the emotional valence score (quantified data) extracted from the analyzed (textual) data to statistically investigate the correlation and effect of the emotions scores (polarization or intensity of words) expressed by the students with their final grade or learning outcome using Sentiment Analysis package and Statistical analysis methods such as the Spearman's rho (¿), Kendall's tau (r), and Kruskal-Wallis H-Test in R. The results shows that while in overall the emotional valence of the students do not influence or determine the outcome of their study or final average grade (p>0.05). When considering the analyzed comments: Pre-and Post-Course; it found that emotional scores (valence) expressed by the students in the Post-Course are more closely related or linked to the final grades than the Pre-Course comments. The paper empirically sheds light on both the pedagogical and socio-Technical implications of the findings and result toward the achievement of higher levels of students learning outcomes or performance and a sustainable educational practice. © 2024 IEEE.