Indian Journal of Science and Technology
Year: 2016, Volume: 9, Issue: 41, Pages: 1-9
F. Ezhil Mary Arasi1*, S. Govindarajan2 and A. Subbarayan2
1Department of Computer Science and Engineering, SRM University, Kattankulathur, Chennai - 603203, Tamil Nadu, India; [email protected] 2Department of Master of Computer Applications, SRM University, Kattankulathur, Chennai - 603203, Tamil Nadu, India; [email protected], [email protected]
*Author for correspondence
F. Ezhil Mary Arasi Department of Computer Science and Engineering, SRM University, Kattankulathur, Chennai - 603203, Tamil Nadu, India; [email protected]
Objectives: In recent times Web Service providers are finding difficult to satisfy the requirements of the consumers due to several reasons. We have made an attempt to identify the specific parameters for evaluation of an efficient functioning of Composite Web Services. The study is based on a data set sampled from a large number of Web Service Providers. Methods/Statistical Analysis: Random sampling technique was adopted for the preparation of the data set. A multivariate statistical technique viz., Factor analysis is used for identifying the factors. Factor analysis method reduces large number of variables into few factors without sacrificing much of explained variability by the variables. Factor analysis begins with the construction of new set of variables based on the relationship in the correlation matrix. Based on principal component analysis we have identified the factor which determines the effective performance of Web Service providers. Findings: The study is based on a data set sampled from a large number of Web Service providers. The results based on the sample data set indicates the components of Web Service Providers Performance factors, Security factors and Trust factors viz., Successability, Best practices and Throughput respectively. These parameters have the higher loadings compared to other extracted QoS parameters. Application/Improvement: The research work was confined to the usage of factor analysis method based on principal component analysis approach. As a future work the other approaches in cluster method viz., centroid method and maximum likelihood method may be attempted to make a comparative study of the extraction of QoS parameters for its effectiveness.
Keywords: Factor Analysis, Principal Components, Quality of Service, Varimax Rotation, Web Service Selection
Subscribe now for latest articles and news.