ModelExplainer is an innovative tool which automatically generates textual summaries of object-oriented data modeling (OODM) diagrams, using information from data modeling tools such as Rational Rose and Visio.

ModelExplainer lets the user focus on any class in a model, and view a detailed description of its relationships to other classes. The hypertext/hypergraphic interface makes it easy to navigate between descriptions of related classes, and several options are available for configuring the generated text.

ModelExplainer can generate text either interactively or in batch mode, for documentation purposes.


  • Instantaneous translation of diagrams to text helps functional analysts to debug data models, by clarifying design aspects such as cardinality and dependence that rely on complex notation when presented graphically.
  • Texts facilitate requirements validation with end users (e.g. domain experts), who may not be familiar with particular data model formalisms or graphical notations.
  • Texts can include automatically generated examples, which often highlight design errors.


  • Lavoie, Benoit; Rambow, Owen; and Reiter, Ehud (1997). Customizable Descriptions of Object-Oriented Models. In Proceedings of the Fifth Conference on Applied Natural Language Processing, Washington, DC, pp. 265-268. [Acrobat, 207 Kb] [PostScript, 1322 Kb]
  • Lavoie, Benoit; Rambow, Owen; and Reiter, Ehud (1996). The ModelExplainer. In Demonstration Notes of the International Natural Language Generation Workshop (INLG-96), Herstmonceux Castle, Sussex, UK, pp. 9-12.