Biography
Dr. Geetha Mary A is a Professor in the School of Computer Science and Engineering (SCOPE) at Vellore Institute of Technology (VIT), Vellore, Tamil Nadu, India. Her primary research area is Data Mining, with broader research interests encompassing databases, intelligent systems, data security, privacy-preserving data mining, and machine learning applications. Dr. Geetha Mary A has been associated with VIT for several years in teaching and research and has contributed to the areas of database systems and data mining. Her earlier academic qualifications include an M.Tech. in Computer Science and Engineering from VIT and a B.E. from the University of Madras. Her research work includes studies in privacy-preserving data mining, anonymization, data deduplication, healthcare data analytics, and intelligent data-processing techniques. She has published research articles in international journals and conferences and has guided research scholars in data-mining-related topics.
Conference contributions 1
A MULTI-MODAL AND EXPLAINABLE AI APPROACH FOR STUDENT RISK ASSESSMENT USING DYNAMIC FUSION
The students of today encounter numerous sources of challenge, including academic demands, emotional fitness, and preparation for their profession. The result is that intelligent support systems need to be provided across a range of integrated interfaces. While there are numerous solutions that address portions of these challenges, they have typically been found to be supported via multiple signatures, and are often located between different platforms that make it difficult for the student to use each service in an efficient manner. This paper will present the Smart Student Suite as an Integrated and AI-enabled student support platform which addresses the four identified areas of support (Academic, Emotion, Career and Writing) as identified. The Academic Module will utilize predictive analytics to identify at risk students and to understand how they are likely to perform in the future. The Emotion Module will utilize multimodal analysis to help determine a student's emotional state and to provide appropriate support in a timely manner. The Writing Module will utilize advanced natural language processing techniques to support the student via support in the improvement of their writing structure, clarity and overall writing performance. The Career Module will also utilize trade fraud detection methodologies to identify unsafe job postings and also to ensure the student's access to legitimate opportunities. By co-locating these features into a single platform, the system is intended to provide a proactive and holistic support environment that will enhance student's academic performance, enhance student's mental well-being, and improve graduate career readiness.