In the year 2011, Han, Jung 265described the Aspect Oriented Programming (AOP) is well suited to clustercomputing software by using simple, intuitive, and reusable aspects. Throughoutqualitative and performance evaluations, AOP significantly improves the codereadability as well as the modularity, and AOP-based software has the sameperformance and scalability as similar software that is developed without usingAOP.
Guabtni, Ranjan 266 concernedwith data provisioning services (information search, retrieval, storage, etc.)dealing with a large and heterogeneous information repository. Increasingly,this class of services is being hosted and delivered through Cloudinfrastructures. Awang 268 proposed an algorithm and analytical model basedon asynchronous approach to improve the response time, throughput, reliabilityand availability in Web Server Cluster. The provision of high reliability inthis model is by imposing a neighbor logical structure on data copies.
Datafrom one server will be replicated to its neighboring server and vice versa inthe face of failures. Senguttuvan, Krishna 283 reviewedfive of the most representative off-line clustering techniques: K-meansclustering, Fuzzy C-means clustering, Mountain clustering, Subtractiveclustering and Extended Shadow Clustering. The techniques were implemented andtested against a medical problem of heart disease diagnosis. Performance andaccuracy of the four techniques are presented and compared.
Soni,Ganatra 284 provided acategorization of some well known clustering algorithms. It also describes theclustering process and overview of the different clustering methods. Bahmani, Moseley 289 proposedinitialization algorithm k – means obtains a nearly optimal solution after a logarithmic number of passes,and then show that in practice a constant number of passes suffices. The “data mining extensions” (DMX) 2 is aSQL-like language for coding data-mining models in the Microsoft platform, andtherefore it is difficult to gain understanding of the data-mining domain. Datamining is a highly complex task which requires a great effort in preprocessingdata under analysis, e.g.
, data exploration, cleansing, and integration9. The 10 provides an entire framework to carry out data mining but, onceagain, they are situated at very low-abstraction level, since they arecode-oriented and they do not contribute to facilitate understanding of thedomain problem. Research papers 11 and 12 provide a modeling framework tode ne data-mining techniques at a high-abstraction level by using UML. However,these UML-based models are mainly used as documentation.
Parsaye 13 examinedthe relationship between OLAP and data mining and proposed an architectureintegrating OLAP and data mining and discussed the need for different levels ofaggregation for data mining.