EMPIRICAL VALIDATION AND ASSESSEMENT OF WEB METRICS

Abstract

Websites have become an integral part of our day to day life. They act as a source of information and also as a medium of communication. With such important characteristic development of web sites should be done carefully. The quality of web sites is typically concerned with performance and usability and is measured using web metrics. Web metrics are the measure of attributes of a web page. Collecting, analyzing and interpreting web metrics is referred as web analytics. In this study we categorize websites into three categories which are collected from pixel awards website and then analyze these categories using web metrics. For analysis we have created a web scrapper tool which evaluates web sites Using web page metrics and applied eight machine learning algorithm such as naïve bayes, bagging, random forest, random tree, multilayer perceptron, nnge and oner on these web page metrics to predict goodness of websites and also classified them within a particular category. The result of this paper will provide an empirical foundation for web site designing.