Analyzing Client-Side Interactions to Determine Reading Behavior

Abstract

Traditional monitoring and user modeling techniques in adaptive hypermedia systems consider pages as atomic units although different sections may refer to different concepts. This has been mainly due to the fact that most user interactions being monitored referred to the request of a new document and there was too little activity information to differentiate between sections of a page. Client-side monitoring can provide additional information on user interactions inside the browser window and may relate them to areas within a document. A user study was carried out to show whether and how this data might be used to identify which parts of a page have been read.