The goal of this workshop is to gather community input and feedbacks, either from industry and academia, on UML consistency rules. This workshop will provide an opportunity for participants who have been working on, or who have a vested interest in UML consistency to interact with each other in a highly interactive venue to consolidate the body of knowledge on UML consistency rules and discuss ideas for further research in this area. It will provide a platform for discussions, interactions and collaborations regarding this topic.
Call for contribution
The format of the workshop will be a mix of short presentations and working group discussions. We are therefore calling for papers that will contribute to those presentations and discussions along one or more of the following topics. (This is one of the three publication outcomes of the workshop.)
Topics of interest
The workshop topics include all those related to UML consistency rules. They will include, but will not be limited to:
Consistency rules for UML diagrams
- Most used diagram: the Class Diagram is the UML diagram most involved in UML consistency rules presented by researchers so far; it is followed in importance by the Interaction Diagram, and the State Machine Diagram. This is not entirely surprising since these are likely the most used UML diagrams. Is the set of rules we have coalesced and that involve those diagrams complete, sufficient?
- Activity Diagrams: according to a recent research, the activity diagram is the second most used UML diagram after the Class Diagram. We however found a very scarce number of rules involving the Activity Diagram. Should the research about UML consistency focus more on the Activity Diagram in the future and what would be additional rules involving the Activity Diagram?
- Package, Profile, Component, Timing, Interaction Overview and Deployment Diagrams: we have not found a single rule involving those diagrams. Should the research about UML consistency focus more on those diagrams in the future and what would be additional rules involving these diagrams?
Dimensions for UML consistency rules
- Vertical consistency and Semantic Consistency: we observed that the great majority of UML consistency rules are Horizontal and Syntactic, and a very few are Vertical and Semantic, while not a single rule was on the Invocation, Observation or Evolution dimensions. Does this suggest that other dimensions than Horizontal and Syntactic are not much relevant to UML or that rules with other dimensions are missing? What would be those missing rules?
Languages to define UML consistency rules, and tool support (for definition and checking)
- The Object Constraint Language (OCL) is the most used language to specify UML consistency rules, followed by Communicating Sequential Processes (CSP) and Promela: using the OCL makes sense since this is a constraint language that is part of the UML specification and used to specify UML’s well-formedness rules. Languages such as CSP and Promela have been used to specify semantic rules between the sequence diagram and the state machine diagram. Other languages less often used to define UML consistency rules are XML Equivalent Transformation, Prolog, Constraint Logic Programming. Is the OCL complete and sufficient to specify UML consistency rules and what would be additional languages that are used nowadays in this field?
- IBM’s tools (e.g., Rational Rose) are the most used tools, followed by Spin, UML/Analyzer and OCL interpreters for checking consistency rules: other tools include Poseidon and ArgoUML. Are the IBM’s tools complete and sufficient to check UML consistency rules and what would be additional tools that are used nowadays in this field?
Consistency rules in UML based domain specific languages
- A UML-based domain specific language extends or adapts the notation and semantics of the UML language: which of the UML consistency rules we have coalesced directly applies to those DSLs? Which other consistency rules do these DSLs need?