DED: Drift Principle in Educational Evolved Data

Clustering data streams is one of the prominent tasks discovering hidden patterns in streams. It refers to process clustering newly arrived into continuously and dynamically changing segmentation patterns. This article presents a stream mining algorithm cluster with focusing on its evolution concept drift. Even though drift expected be present streams, explicit detection rarely done algorithms. Concept caused by changes distribution over time. Relationship between occurrence physical events has been studied applying education stream. Viber produced Groups our Computer Science Department are used conduct this study. The results show that proposed superiority existing ones purity, entropy, sum square error measurements. Experiments led conclusion accompanied change number clusters outliers indicates significant event. kind online monitoring can utilized systems various ways, such as capabilities participants.