What's in a Session: Tracking Individual Behavior on the Web

TL;DR

This study analyzes detailed web browsing data from undergraduates to reveal properties of web traffic, user behavior, and session definitions, challenging traditional session segmentation methods and proposing a logical, referrer-based approach.

Contribution

It introduces a new logical definition of web sessions based on referrer URLs and demonstrates its advantages over timeout-based methods.

Findings

Web site popularity distribution can be unbounded and lacks a mean.

Individual browsing behavior often follows a log-normal distribution.

Timeout-based session segmentation significantly alters browsing statistics.

Abstract

We examine the properties of all HTTP requests generated by a thousand undergraduates over a span of two months. Preserving user identity in the data set allows us to discover novel properties of Web traffic that directly affect models of hypertext navigation. We find that the popularity of Web sites -- the number of users who contribute to their traffic -- lacks any intrinsic mean and may be unbounded. Further, many aspects of the browsing behavior of individual users can be approximated by log-normal distributions even though their aggregate behavior is scale-free. Finally, we show that users' click streams cannot be cleanly segmented into sessions using timeouts, affecting any attempt to model hypertext navigation using statistics of individual sessions. We propose a strictly logical definition of sessions based on browsing activity as revealed by referrer URLs; a user may have several active sessions in their click stream at any one time. We demonstrate that applying a timeout to these logical sessions affects their statistics to a lesser extent than a purely timeout-based mechanism.