AI-DRIVEN ANALYSIS OF SURVEY RESPONSES ON ACCESSIBILITY SUPPORT IN HIGHER EDUCATION INSTITUTIONS

This study employs machine learning (ML) techniques to analyze open-ended survey responses collected from higher education institutions (HEIs) across the European Union, with a specific focus on accessibility support for visually impaired students. Utilizing clustering, topic modeling, and sentiment analysis, the research identifies recurring themes, evaluates institutional attitudes, and reveals patterns in accessibility-related policies, support mechanisms, and emerging trends, particularly the integration of artificial intelligence (AI). Sentiment analysis, implemented through natural language processing (NLP), serves as a robust methodological approach to extracting meaningful insights from qualitative data. Its application in this study demonstrates the potential to quantify subjective experiences and perceptions captured through open-ended survey questions, thus enriching the overall analysis of student-centered accessibility initiatives. In parallel, factor analysis was conducted on structured survey items to further examine institutional support mechanisms. This analysis revealed three principal factors: institutional support structures, the existence and quality of accessibility policies, and the physical accessibility of campus infrastructure. These factors highlight both strengths and gaps in the provision of services for visually impaired students within the participating HEIs. The study’s findings indicate generally positive institutional sentiment toward accessibility, while also identifying notable deficiencies in resource allocation and policy implementation. Furthermore, the results underscore the potential of emerging technologies, particularly AI-based solutions, to enhance accessibility and inclusivity in higher education settings. The research was conducted as part of a broader initiative at Georgian Technical University, contributing to the understanding of how European HEIs address the educational needs of visually impaired students and offering actionable insights for future policy and technological development.