Leveraging Text Mining for Analyzing Students' Preferences in Computer Science and Language Courses

Abstract

In an increasingly competitive, globalized world, educational institutions must strategically offer courses that align with the skill-acquisition needs of their target audience. As such, the application of text mining techniques to extract valuable insights and patterns from structured data across various knowledge domains becomes paramount. This study employed text mining to scrutinize students' preferences for course offerings at the Computer and Language Center of the National University of Jaen. The analysis was based on data collected from a Google Forms survey of 315 students. The employed methodology facilitated the unearthing of patterns, trends, and semantic relationships within a large corpus of students' opinions. Frequency distributions and word clouds were generated using R programming language. Furthermore, the WEKA software and Python were utilized for cluster analysis, enabling the detection of groupings and trends within the data. Although other methods such as sentiment analysis and statistical methodologies exist, text mining was deemed most suitable for identifying patterns and relationships within students' opinions. The study revealed that students predominantly favored advanced Excel, AutoCAD, ArcGIS, Nutrition, and Revit courses, which appeared to correlate with their professional aspirations and prevailing course trends. Therefore, the application of text mining tools to analyze structured institutional data can significantly contribute to informed decision-making processes.