Non-Monotonic Knowledge Evolution in VLKDBs

Non-monotonic knowledge evolutions and exceptions constitute a complex theoritical and practical problem. The state of the art shows a rich and surprising diversity of approaches. Their study along a common framework reveals a few basic issues which require further theoritical investigation. In the context of very large knowledge/data bases (VLKDBs), specific constraints must be taken into account. They require a specific approach of the problem, which can be sketched by a set of basic guidelines. In order to illustrate the importance of the problem, a proposal called semantic tolerance is briefly described. Its various relationships with user interfaces, rule administration, advanced design techniques and long-term transactions show that non-monotonic knowledge evolution constitute a central problem in advanced databases, and consequently would deserve a major research effort by the database community.

• Type: Conference
• Year: 1990
• Where: VLDB
• Authors: Christian Esculier
• Keywords: Database, Large Knowledge/data Bases, Non-monotonic Knowledge Evolution, Practical Problem, VLDB 1990