CLUSTERING MULTI ATTRIBUTE SIMILARITY INDEX FOR CATEGORICAL DATA STREAMS

Deena Babu Mandru
2018 International Journal of Advanced Research in Computer Science  
Data mining is an aggressively concept in information retrieval based on different attributes from different data sources. For effective data collection from data sources with respect to relevant data, one-class learning is required to perform labeled based classification with individual training sequences on attributes. In clustering, uncertain data with different data set visualization. Uncertain One Class Clustering (UOCC) with support vector machine to explore data summarization in terms of
more » ... user preference. UOCC process single attributes from reliable data streams for inconsistent data. So that in this paper we propose Clustering with Multi-Attribute Framework (CMAF) to group multiple attributes to explore uncertain data from reliable data. CMAF construct matrix with different reliable attributes based on relevant features. Proposed approach defines effective data summarization for relevant data with attributes partitioning and constructs user profile based on relative attributes. Experimental results come out for proposed approach gives better and expressive results with comparison of state of art methods.
doi:10.26483/ijarcs.v9i3.5931 fatcat:rfy3jtz5bzbh7ajsfdplib7wc4