The CSB approach to prediction of chemical reactions

The methodology and recent advances in developing the chemical sense builder (CSB) system for simulation of organic reactions are presented. This system comprises two functional modules that can be used separately or in combination. Four logic-based and knowledge-based models for reaction generation and discovering constitute the first module. The second one, newly designed, provides knowledge acquisition and learning tools for exploration and derivation of knowledge that can be employed in the reaction simulation process. An overview of the CSB programming tools and a knowledge source are given. The new CSB features are illustrated by an example concerned with the generation and evaluation of example reactions. (C) 2004 Elsevier B.V. All rights reserved.