A flexible graph-unification formalism and its application to natural-language processing

A graph-unification-based representation language is described that was developed as the grammar formalism for the LILOG research project at IBM Germany. The Stuttgart Unification Formalism (STUF) differs from its predecessors in its higher flexibility and its algebraic structure. It is well suited for the implementation of rather different linguistic approaches, but is currently employed mainly in the development of Categorial Unification Grammars with a lexicalized compositional semantics. Examples from the syntactic and semantic processing of natural language are used to illustrate the virtues of the formalism and of our lexicalist approach to linguistic analysis.