Probabilistic Reasoning in DL-Lite
Author(s)
Ramachandran, R
Qi, G
Wang, K
Wang, J
Thornton, J
Year published
2012
Metadata
Show full item recordAbstract
The problem of extending description logics with uncertainty has received significant attention in recent years. In this paper, we investigate a probabilistic extension of DL-Lite, a family of tractable description logics. We first present a new probabilistic semantics for terminological knowledge bases based on the notion of types. The semantics proposed is not capable of handling assertional knowledge. In order to reason with both terminological and assertional probabilistic knowledge, we propose a probabilistic semantics based on a finite semantics for DL-Lite called features. This approach enables us to infer new ...
View more >The problem of extending description logics with uncertainty has received significant attention in recent years. In this paper, we investigate a probabilistic extension of DL-Lite, a family of tractable description logics. We first present a new probabilistic semantics for terminological knowledge bases based on the notion of types. The semantics proposed is not capable of handling assertional knowledge. In order to reason with both terminological and assertional probabilistic knowledge, we propose a probabilistic semantics based on a finite semantics for DL-Lite called features. This approach enables us to infer new information from the existing knowledge base by drawing on the inherent relation between a probabilistic TBox and a probabilistic ABox.
View less >
View more >The problem of extending description logics with uncertainty has received significant attention in recent years. In this paper, we investigate a probabilistic extension of DL-Lite, a family of tractable description logics. We first present a new probabilistic semantics for terminological knowledge bases based on the notion of types. The semantics proposed is not capable of handling assertional knowledge. In order to reason with both terminological and assertional probabilistic knowledge, we propose a probabilistic semantics based on a finite semantics for DL-Lite called features. This approach enables us to infer new information from the existing knowledge base by drawing on the inherent relation between a probabilistic TBox and a probabilistic ABox.
View less >
Conference Title
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume
7458 LNAI
Subject
Artificial intelligence not elsewhere classified
Information and computing sciences