A More Perfect Union: Enabling Collaborative Goal Modeling through Model Merge

A More Perfect Union: Enabling Collaborative Goal Modeling through Model Merge
Kathleen Hablutzel, Anisha Jain

During goal modeling elicitation, stakeholders may create goal models reflecting their individual perspectives. We proposed an automatic merging algorithm to combine two or more goal models for increased analysis precision and the resolution of conflicting information. Our algorithm merges intentions, actors, these elements’ respective relations, and evolving function and timeline information using the principles of gullibility, contradiction, and consensus.

Our semi automatic approach requires little user intervention and runs under 10 ms on a conventional laptop. It is also scalable, quickly merging models with large numbers of elements, and is associative, with n-way models merging to the same model regardless of order. Inspectors validated our approach by hand-merging models and comparing them with auto-merged models and found that the results were comparable, as the majority of models were rated as mostly to almost totally complete and correct. For future work we intend to implement an automatic layout algorithm, explore state of the art matching algorithms, and extend our validation.