Decision support based on time-series analytics: A cluster methodology

Abstract

Web analytic techniques have become increasingly popular, particularly Google Analytics time-series dashboards. But interpretations of a website's visits traffic data may be oversimplified and limited by Google Analytics existing functionalities. This means website mangers have to make estimations rather than mathematically informed decisions. In order to gain a more precise view of longitudinal website visits traffic data, the researchers mathematically transformed the existing Goggle Analytics' log data allowing the vectors of website visits per each year to be considered simultaneously. The methodology groups the data of an example website gathered over an 'x' year period into 'y' clusters of data. The results show that the transformed data is richer, more accurate and informative, potentially allowing website managers to make more informed decisions concerning promoting, developing, and maintaining their websites rather than relying on estimations. © 2014 Springer International Publishing.