Context-Aware Generation and Adaptation of workflows (PhD Thesis, finished)

Workflows have to be flexible in order to react in a best possible way to individual user requirements, dynamic changes in time and location, unforeseen events, or necessary business process changing.

Our example scenario is a climbing guide which customizes the proposals for climbing areas, implemented as workflows, according to the user’s profile and the dynamic changes of the user’s context as time, location or weather. As context in this application, we consider for instance the user preferences (climbing grade 7, likes to go outside for climbing even when the temperature is low), location, time, and weather.

In general, context-aware adaptations should address only these parts of the process depending on external changes while not touching – if possible – the others. Thus, the goal is to change a process without complete replacement.

There are two options to reach workflow flexibility:

Our goal is to address the dynamic and automatic customization of the workflow specification according to the surrounding context. The idea is not to have a complete workflow specification but generic templates with placeholders. The placeholders are replaced by sub workflows according to the current context (user preferences and dynamic context) to generate a complete specification.

While the specification template of both workflow instances is the same, the generated workflow types are different because of different context information. While the workflow 1 is not able to access to profile information the workflow has to analyse context information for outdoor and indoor climbing (opening hours and weather) and the proposed list of climbing areas is larger than for workflow 2. Workflow 2 offers a more accurate result list as it has access to the profile information and thus is able to customize (filter) the result list to climbing halls only.

Staff

Research Topics