The project involves accumulation of user data from Google Analytics through its API, classification of data using a naive Bayesian filter, and visualisation of data with JPGraph.

Situation/ Problem

The client with an online sales portal was trying to find the trend of users on their public sales pages. The client was interested to know the rate of bounce and patterns of user interaction on the sales pages. The procedure was complex due to the high-volume of data gathered.


To start with we reviewed the Google Analytics pages to understand the trend of traffic. We found out that data collected at Google Analytics are adequate and we do not have to go back to Apache log. We then retrieved the Analytics data to a MySQL database. We analyzed the data on time-series, but understood that it requires statistical analysis of data. We employed a naive Bayesian filter for this and visualized the results with the help of JPGraph.


The client was able to take bold decision about reorganising their online sales process user interface to boost their sales.