4 November 2009 14:15 in CM221
Since the advent of the Internet, the ability of websites to track the visitors has been considered one of the most promising facets of the new media. Clickstream data gathered from a web site can provide insight into the behavior, buying habits and preferences of the website visitors who can be considered as prospective customers. Detailes of Web usage behaviour provide researchers or Web managers the opportunity to study how users browse or navigate thorough websites and to assess site performance in various ways. For this purpose, we use data from commercial websites belonging to clients of a local web management company. These websites, selling products and services on the Internet. Server log files provide tracking data which contain the general clickstream information from the website visitors. We also have the conversion data which comprises the converted visitor information such time, date, IP, agent, and amount of conversion. In this session we will present some explanatory data analysis performed over the clickstream data set. This involves data visualizations, fitting statistical models to predict probablistic behaviour of the variables, as well as descriptive modeling by logistic regression to describe conversion behaviour.
See the Stats4Grads page for more details about this series.