Algorithm for Adapting Cases Represented in a Tractable Description Logic
This work addresses a specific bottleneck in knowledge representation for AI systems using description logics, providing a foundational tool for CBR applications in domains like semantic web or medical diagnosis.
The paper tackles the challenge of case adaptation in description logic-based case-based reasoning by extending the logical basis from propositional logic to the tractable DL $\mathcal{EL_{ot}}$, presenting a syntax-independent and fine-grained algorithm for this formalism.
Case-based reasoning (CBR) based on description logics (DLs) has gained a lot of attention lately. Adaptation is a basic task in the CBR inference that can be modeled as the knowledge base revision problem and solved in propositional logic. However, in DLs, it is still a challenge problem since existing revision operators only work well for strictly restricted DLs of the \emph{DL-Lite} family, and it is difficult to design a revision algorithm which is syntax-independent and fine-grained. In this paper, we present a new method for adaptation based on the DL $\mathcal{EL_{\bot}}$. Following the idea of adaptation as revision, we firstly extend the logical basis for describing cases from propositional logic to the DL $\mathcal{EL_{\bot}}$, and present a formalism for adaptation based on $\mathcal{EL_{\bot}}$. Then we present an adaptation algorithm for this formalism and demonstrate that our algorithm is syntax-independent and fine-grained. Our work provides a logical basis for adaptation in CBR systems where cases and domain knowledge are described by the tractable DL $\mathcal{EL_{\bot}}$.