Briefing - Interaction with semantically enhanced content
Any specific use case is associated with an APPLICATION DOMAIN.
An APPLICATION DOMAIN usually has a THEMATIC SCOPE within the world of all possible WEB RESOURCES.
An APPLICATION DOMAIN can be characterised by DOMAIN TASKS, USER ROLES, and (specific) CONTENT ARTEFACTS.
CONTENT ARTEFACTS can presented in specific ways - the result is a PRESENTATION ARTEFACT.
If it is foreseen that the end user should be able to further interact with the PRESENTATION, then that INTERACTION needs to be specified.
In a knowledge-based content application, the INTERACTION should be related to the APPLICATION DOMAIN of the end user.
Aims of the workshop are:
- A Taxonomy of Interaction Patterns
- support for very frequent Presentation Artefacts, e.g. tables and lists
- Create a list of interaction patterns supported by IKS, design and implement two interaction patterns
Technical agreements were:
- T5.1 should deliver basic interaction widgets (executable code) that support the construction of complex user interactions ("domain tasks")
- where the number of possible widgets exceeds project funds, design specifications for further widgets should point external developers in the right direction
- where designs may vary significantly, a methodology should be provided, how to achieve the target interaction irrespective of implementation technology
The Interaction patterns that are being implemented should satisfy the following constraints: - be motivated through one of the use cases - be exemplified (i.e. used) in at least two of the use cases - be documented and placed in the taxonomy developed at http://wiki.iks-project.eu/index.php/InteractionPatterns
We also discussed possible ways of specifying control over models and views which are inherent in the interaction patterns. This will be needed to define a methodology for building IKS-type knowledge and content rich applications. There was some discussion about the relative merits of SCXML or jBPM and similar notations for expressing process models. It remained unclear whether significant support for process/task/workflow engines is a useful objective for IKS. The agreement is that down-to-earth, reusable interaction components (widgets) would have most impact on the CMS sector.
ToDo: Extend the list of interaction patterns (DFKI) and follow a preliminary template: Problem description, Solution, Usage Examples, Widgets and Restful Services, link to the code library (ui-widgets)
Participants: Tilman Becker, Massimo Romanelli, Stefan Germesin (DFKI); Wernher Behrendt, Rupert Westenthaler, Violeta Damjanovic, Szaby Grünwald (Salzburg Research)
Location and Time: Salzburg Research, Nov 2-3, 2010