On Categorial Grammatical Inference and Logical Information Systems
Abstract
In this chapter, learning is viewed as a symbolic issue in an unsupervised setting,from raw or from structured data, for some variants of Lambek grammars and of categorial dependency grammars. For these frameworks, the authors present different type connectives and structures, some limitations, and some algorithms. On the experimental side, categorial grammar has potentials as a particular case of Logical Information System.