Most of the current WWW is made up of dynamic pages. The development of dynamic pages is a difficult and costly endeavour, out-of-reach for most users, experts, and content producers. We have developed a set of techniques to support the edition of dynamic web pages in a WYSIWYG environment. In this paper we focus on specific techniques for inferring changes to page generation procedures from users actions on examples of the pages generated by these procedures. More specifically, we propose techniques for detecting iteration patterns in users’ behavior in web page editing tasks involving page structures like lists, tables and other iterative HTML constructs. Such patterns are used in our authoring tool, DESK, where a specialized assistant, DESK-A, detects iteration patterns and generates, using Programming by Example, a programmatic representation of the user’s actions. Iteration patterns help obtain a more detailed characterization of users’ intent, based on user monitoring techniques, that is put in relation to application knowledge automatically extracted by our system from HTML pages. DESK-A relieves end-users from having to learn programming and specification languages for editing dynamic-generated web pages.

The work reported in this paper is being supported by the Spanish Ministry of Science and Technology (MCyT), project number TIC2002-1948

Revised Selected Papers of the Joint Working Conferences EHCI-DSVIS 2004, Hamburg, Germany, July 11-13, 2004

The final publication is available at Springer via http://dx.doi.org/10.1007/11431879_10