Proposed Approach for Analysis and Visualization of Educational Data, Based on the Concept of Big Data

This paper introduces a framework for data analytics in a blended learning environment, based on Big Data technologies.The framework integrates heterogeneous educational data sources and enables real-time processing using Apache Spark.The methodology applies Z-score and Min-Max normalization and uses Principal Component Analysis (PCA) for dimensionality reduction.K-means clustering is employed to identify patterns in student behavior.The comparison of normalization methods shows that Min-Max normalization produces more compact clusters than Z-score.The analysis also indicates consistent relationships between students' activity on the Moodle platform and their academic outcomes.The study contributes a Spark-based procedure for descriptive, diagnostic, and cluster analytics.It also offers an empirical evaluation of normalization methods.In addition, it provides visual analytics that support early identification of at-risk students.These tools can help teachers improve course organization.