Jérôme Cardot, Université Paris IV Sorbonne
LDO provides a way of representing objects as built on concepts by two operators: an object builder and an object determinator. Objects built by help of these operators are more or less determinate, LDO giving a model of categorization; objects may accept several descriptions, so that recognizing different descriptions of a same object is actually an important task.
In order to compute identification of these objects, this paper introduces formal languages to represent them, so that they can be processed with either classical (automata, congruence) or newer (like S-languages) computer science methods. LDO needs this to model objects of various areas, such as lexical representation, NLP, OO programming...