Presenting Herbrand Models with Linguistically Motivated Techniques

Model generation refers to the automatic construction of models for first-order logical theories and is used for instance to find solutions to logically encoded problem specifications. Handling these results may require an inspection of the generated models for crucial information, which is insufficiently supported by current simplistic presentation techniques. Improving these methods considerably, we adopt a number of linguistically motivated techniques for this purpose, including filtering out assertions considered inferable by the addressee, and aggregation of assertions sharing information. Through the incorporation of these techniques, presentations focus on interesting portions and differences across models, especially supporting the discovery of flaws in problem specifications.