Implementation of Classification Techniques for Finding the Relevance of Fields and Supervisors in Final Projects

This study aims to analyze the alignment between research fields and supervisors in the final project documents of Information Systems students. This descriptive quantitative research with survey method. The sample used of student final project documents. The sample was selected using the purposive sampling technique, focusing on Information Systems students' final projects from 2021–2023. A validation sheet was used as the research instrument to assess the alignment between fields and supervisors. The research procedure began with data labeling of fields of expertise and supervisors, conducted by two expert lecturers. To ascertain the precision of the alignment between research domains and supervisors, the data was further subjected to classification approaches backed by the Naïve Bayes Classifier, Decision Tree, Support Vector Machine (SVM), Logistic Regression, Random Forest, and K-Nearest Neighbor (KNN) algorithms. The algorithm results showed that the alignment between research fields in the final projects and supervisors' expertise ranged from 50-70% accuracy. The algorithm performance review indicated that the Random Forest algorithm was the best for both types of data, with an accuracy of 66.66% (supervisors) and 73.00% (fields). Further test results indicated that Random Forest is effective in predicting the alignment between fields and supervisors for final project titles.