Challenges and Future Research Directions for Stream Clustering

Since the incoming data is unlabeled, streams are more closely tied to clustering. Researchers have focused on stream analysis since streams are produced by a variety of stream applications, including cell phones, GPS, autos, security sensors, televisions, social media, and more. An extensive survey is conducted on the problems relating to data stream clustering in this study. This study discusses various forms of drift, windowing techniques, drift detection technique and clustering algorithm that have been recently employed in the literature to quickly handle incoming data, and an open research questions are conducted and potential future research areas for big data stream clustering.