Integrating AI-Based Sentiment Analytics and Emotional Intelligence to Predict Academic Achievement among Secondary School Students

ABSTRACTThe growing use of Artificial Intelligence (AI) in education has made it possible to look athow students learn by using data-driven methods like sentiment and learning analytics. AI-based analytics can forecast academic outcomes by detecting engagement patterns;nevertheless, students' emotional capabilities may affect the efficacy of these insights inenhancing performance. The current study investigated the predictive influence of AI-drivensentiment and learning analytics on academic performance among secondary school pupils,with emotional intelligence regarded as a mediating variable.A quantitative descriptive correlational design was utilized. The sample comprised 200students from Classes IX and X, chosen via random sampling. AI-generated sentiment andlearning analytics metrics were extracted from structured five-point learning experienceratings that indicated interest, engagement, and feedback response in AI-enhanced scienceeducation settings. A standardized scale was used to measure emotional intelligence, andannual exam scores were used to measure academic success. Descriptive statistics, Pearson'scorrelation, and regression-based mediation analysis were used to look at the data.The results indicated substantial favourable correlations between AI-derived sentimentindicators, emotional intelligence, and academic performance. Emotional intelligencepartially mediated the association between AI-derived indicators and academic achievement,indicating that emotional competences affect students' responses to analytics-informedlearning feedback. The research underscores the significance of amalgamating emotional skillenhancement with AI-driven learning analytics to foster equitable, human-centric educationalmethodologies.