A Comparative Sentiment Analysis of Computer Engineering Student Feedback Using Decision Trees and SVM

The University of Borneo Tarakan, like many Indonesian universities, is committed to continuous quality improvement in education services. A crucial aspect of this improvement is gathering and analyzing student feedback to enhance lecturer performance. This research focuses on analyzing student comments using sentiment analysis, a technique that categorizes text into positive, negative, and neutral sentiments. To achieve this, two machine learning algorithms were employed: Decision Trees and Support Vector Machines (SVM). The research involved two approaches: Lexicon-Based Sentiment Analysis and TF-IDF word weighting. The Lexicon-Based approach compared the automated sentiment classification with manual human categorization to assess accuracy. The TF-IDF method, on the other hand, aimed to improve classification accuracy by assigning weights to words based on their frequency and importance. The experimental results demonstrated that Decision Trees outperformed SVM in terms of classification accuracy, achieving 95.454546% compared to 94.805194%. This finding suggests that Decision Trees is a more effective technique for sentiment analysis of student comments in this specific