Computational Model For Classification Rule Mining Using Large Dataset

Abstract

In this thesis documents classification has been discussed for large data sets. In classification process improvement of accuracy and efficiency are very challenging task for researchers, specially during handling large data set. In this research work a framework for knowledge extraction using document classification has developed for large data sets and gives better classification accuracy and efficiency. newlineIn present research work, four models for document classifications are proposed. newline1. Computational model for document density based classification. newline2. Document Classification for Large Datasets Based On Hesitant Fuzzy Linguistic Term Set. newline3. Hamming k-Nearest Neighbor (Hk-NN) Classifier for Document Classification newline4. Normalized hamming k-Nearest Neighbor (NHk-NN) Classifier for Document Classification. newlineThe first proposed approach work based on document density and classify the document efficiently as compare to existing methods. The percentage of classification accuracy is improved. newlineThe second proposed approach is a hesitant fuzzy linguistic term set based classifier. It classify the document very fast over the large data sets compare to the existing methods likes k- Nearest Neighbor, Centroid-based and SVM. newlineThe third proposed approach is a hamming k-Nearest Neighbor (Hk-NN) Classifier for Document Classification. It works based on hamming distance shown better performance over the distance based approches. newline