Fuzzy sets in Approximate Reasoning and Information Systems
Résumé
The representation of human-originated information and the formalization of some forms of commonsense reasoning motivated the creation and development of fuzzy sets. Yet there is still a long-lasting misunderstanding about the relationship between fuzzy logic and subfields of computer science such as Artificial Intelligence (AI) and Information Systems Technology (IST). AI and 1ST have been dominated by a tradition that puts emphasis on symbolic information and neglecting imprecision as well as uncertainty. What fuzzy sets typically bring to these fields is a mathematical framework for capturing gradedness in information representation and reasoning devices. As a consequence, the fuzzy approach is somewhat at odds with the AI tradition, because it suggests that there is more to AI and 1ST than symbolic processing. Indeed, fuzzy logic offers a flexible interface beween symbolic and numerical information. Of course, the gradedness of membership functions of fuzzy predicates can be variously interpreted: similarity between propositions, levels of uncertainty, and degrees of preference, all of which play a role in human reasoning. This book provides an account of the currently available concepts and methods from fuzzy set and possibility theory in the fields of knowledge representation, automated reasoning, knowledge-based systems, learning, data fusion, and information systems.