Akademik Veri Yönetim Sistemi
Bitlis Eren Üniversitesi Fen Bilimleri Dergisi, cilt.15, sa.1, ss.188-200, 2026 (TRDizin)
Exploratory Data Analysis (EDA), recognized as the initial and most critical phase of the data science workflow, plays a fundamental role in understanding the structure of datasets, performing data cleaning, and preparing data for subsequent modeling tasks. This study introduces an interactive EDA platform developed with the R programming language and the Shiny framework. The platform allows users to upload datasets and conduct essential statistical analyses and visualizations, while additionally incorporating large language models (LLMs), such as the OpenAI GPT-4-turbo model, to automatically generate explanatory insights and interpretative commentary regarding the data. By complementing traditional statistical evaluations with language model–driven perspectives, the proposed approach enriches the analytical process by enhancing user intuition and interpretive depth. The system was evaluated using sample datasets, through which both conventional EDA outputs and LLM-assisted interpretations were demonstrated. The findings suggest that the integration of LLMs within Shiny applications holds considerable potential to advance data science education, decision support systems, and automated reporting practices.