Abstract
This research introduces a technique for using an unsupervised machine learning method to data mine cultural datasets on the cultural information system (CIS). This has been referred to as web data mining for cultural data analytics and it will help web users analyze large, historical cultural dataset objects within web clusters. However, current techniques are ineffective at classifying data points into specific CIS clusters. This issue was resolved by using a Hybridized Single Sliding Centroid with Density-based Clustering (HSSCDC) to optimize data analytic techniques. Density-based clustering improves the clustering precision of the cultural analysis system (CAS) and optimizes window size. Experimental validation shows that the proposed HSSCDC automatically identifies data patterns in CAS web clusters with greater precision, less computational complexity, and less error than previous methods. This will help web users predict historical dataset objects more accurately and minimize computational time.