Fragmenting very large XML data warehouses via K-means clustering algorithm

Abstract

XML data sources are gaining popularity in the context of Business Intelligence and On-Line Analytical Processing (OLAP) applications, due to the amenities of XML in representing and managing complex and heterogeneous data. However, XML-native database systems currently suffer from limited performance, both in terms of volumes of manageable data and query response time. Therefore, recent research efforts are focusing on horizontal fragmentation techniques, which are able to overcome the above limitations. However, classical fragmentation algorithms are not suitable to control the number of originated fragments, which instead plays a critical role in data warehouses. In this paper, we propose the use of the K-means clustering algorithm for effectively and efficiently supporting the fragmentation of very large XML data warehouses. We complement our analytical contribution with a comprehensive experimental assessment where we compare the efficiency of our proposal against existing fragmentation algorithms.

Cloud-Based Machine Learning Tools for Enhanced Big Data Applications

“…An obvious extension of the work described in this paper is to use other benchmarks in order to include other workload activities. Also, we plan to further improve the characteristics of our framework by integrating solutions for dealing with novel aspects of massive big data set processing, on top of which workloads may be still defined, such as data compression techniques (e.g., [15]), fragmentation approaches (e.g., [13]), privacypreservation approaches (e.g., [14]) that, particularly, may be extremely useful combined with malware detection issues.…”

Coarse-grained workload categorization in virtual environments using the Dempster-Shafer fusion

“…An obvious extension of the work described in this paper is to use other benchmarks in order to include other workload activities. Also, we plan to further improve the characteristics of our framework by integrating solutions for dealing with novel aspects of massive data set processing, on top of which workloads may be still defined, such as data compression techniques (e.g., [14]),jragmentation approaches (e.g., [12]), privacy-preservation approaches (e.g., [13]) that, particularly, may be extremely useful combined with malware detection issues.…”

Runtime Anomaly Detection in Embedded Systems by Binary Tracing and Hidden Markov Models

“…In this case, our approach can be extended as to capture this important scenario as well, e.g. by exploiting data fragmentation techniques in distributed environments (e.g., [4], [5], [7]), as to improve performance. APPENDIX CPU2006 SPEC benchmark.…”