An Improvised Topsis Approach to Select Web Source as External Data Source for Web Warehousing

Abstract

Objective: The main objective of the paper to incorporate the external web-data efficiently to web-warehouse, as the evolution of web and the requisite of data analytics necessitate it for effective decision support system. Methods/Statistical Analysis: Since the data owned of any organization is insufficient for decision support system. Nevertheless dynamic and complex nature of web pose various challenges during selection of relevant web-data. So evaluation of web resources to select as external source for web-warehouse is the crucial phase during warehousing. Various Multi Criteria Decision Making (MCDM) approaches have been used for it. All these approaches evaluate the web resources on the basis of a set of features which define the relevancy of the resource. Findings: The main focus is on one of the approaches of MCDM viz. Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) approach and also improvised the TOPSIS approach for efficient evaluation of the web resources. In traditional TOPSIS approach Euclidean distance has been measured to compute the proximity of real web-sources from Ideal web-sources. The Euclidean distance measure only the distances between the real and ideal web-resources but not the differences between them. In order to compute the differences between real and ideal web-resources Kullback-Leibler divergence method has been incorporated in the place of Euclidean distance method. Application/Improvements: The improvised TOPSIS computes symmetric as well as asymmetric distances to compute the differences, so efficient to compute the proximity in order to evaluation of web-resources.