chinna1986 / visitante Goto Github PK

Introduction
============
The goal of visitante is to calculate web analytic metric as defined by 
Avinash Kaushik (http://www.kaushik.net/avinash/) on the Hadoop platform


Blogs
=====
The following blogs of mine are good source of details of visitante

http://pkghosh.wordpress.com/2012/06/05/big-web-analytic/
http://pkghosh.wordpress.com/2012/08/10/big-web-checkout-abandonment/


Hadoop Jobs
===========
Set of Hadoop based tools for web analytis. Currently includes the following

- SessionExtractor map reduce which processes w3C compliant web log files and outputs
	- sessionID
	- userID
	- session start time
	- page visited
	- time spent on each page visited

- SessionSummarizer map reduce which processes w3C compliant web log files and outputs
  It provides imprtanr metrics like bounce rate, page depth, abandoned checkout
	- sessionID
	- userID
	- num of pages visited
	- total time spent in session
	- last page visited in session
	- flow status (e.g., whether checkout flow was entered, entered but not completed or completed)
	
- Bayesian discriminant analysis for visitor conversion prediction

Will add the following

- Extend bayesian discriminant analysis for multiple input parameters