Description
This book is a revised version of the PhD dissertation written by the author at the University of Seville, Spain.
It proposes a task mining framework that enriches user interface (UI) logs by incorporating visual information through screenshots and eye-tracking data. This approach enables task mining in restrictive industrial environments, such as the secured or virtualized connections common in outsourcing, overcoming the technical limitations of traditional loggers. Beyond merely identifying activities, the framework enables the extraction of underlying decision models to accurately capture the 'as-is' execution of tasks. Consequently, it allows for the construction of decision trees that explain user choices in greater depth, significantly accelerating the analysis of processes targeted for automation. The proposed framework has been validated through a case study involving synthetic mockups and real-life screenshots, demonstrating both a high level of accuracy in capturing user decisions and its practical usefulness in real-world automation initiatives.
1 Introduction.- 2 Background.- 3 Related Work.- 4 Enriched Behavior Monitoring.- 5 Screen-based Task Mining.- 6 Support Tools.- 7 Validation.- 8 Contributions, Future Work and Conclusions.
