Exploration of Knowledge Bases Inspired by Rough Set Theory

In this paper, we discuss some issues related to exploration of knowledge bases inspired by the rough set theory, which emerged 30 years ago, and is nowadays a rapidly developing branch of artificial intelligence. The partition of rules approach allows us to divide a large set of rules into smaller subsets that are easier to manage. Optimisation relies on reducing the number of rules searched in each run of inference. The paper presents the definition of the knowledge base model based on partition of rules and a modification of the forward inference algorithm for groups of rules generated by the partition strategy. It also contains a simple case study with an example of the partition of rules and the inference process for a simple knowledge base as well as experimental results.