Modeling web user surfing behavior using information scent

Researchers have been making huge strides understanding how users surf for information on the web. Ed Chi and the premier researchers at the Palo Alto Research Center (formerly Xerox PARC) have been building a model of how users seek information when presented with a page of content. Using this model, called Information Scent, PARC researchers have globally modeled user traffic patterns on the web.

In this talk, Ed Chi will describe PARC's groundbreaking InfoScent™ project, specifically pointing out research systems that might help you design a better web site. According to many Web usability experts, the #1 reason for bad web site usability is navigation problems. Ed will share the latest research on automatic tools to assess user behavioral patterns—and help you solve your site's navigation problems.

Ed will describe InfoScent™ Simulator, an experimental service that automatically analyzes the information cues on the Web site to determine probability of user task success. He will also describe the prototype InfoScent™ Analyzer, a complete system for finding user traffic composition by directly inferring users' goals from server logs. more info