Sentiment analysis of prospective mathematics teacher on reasoning and proof questions using Naïve-Bayes classification

Abstract

Reasoning and proof is the ability to arrange patterns, make conjectures, test conjectures, and carry out logical proof, which can be seen as negative or positive for students when solving proof problems requiring reasoning skills. Prospective mathematics teachers in the Mathematics Education Study Program of FKIP Universitas Sriwijaya became the subject of this study. As the research instrument, the questionnaire was designed to explore students' opinions and responses to problems with significant logical difficulty. Sentiment analysis was used to analyze the opinions of prospective mathematics teachers on reasoning and proof questions by grouping positive and negative opinions. The technique used in this study uses the Naïve Bayes algorithm. The classification results in this study were 58.2% positive and 41.86.7% negative, with a total of 70 data. The final result achieved an accuracy value of 53.33%, signifying the reliability of the Naïve Bayes algorithm in understanding and classifying the complex spectrum of sentiments expressed by students. The implications of these findings go beyond sentiment analysis, providing valuable insights for educators, curriculum developers, and policymakers in designing learning strategies and educational policies that can improve mathematics students' reasoning and proof abilities.