The representation of uncertainty in the semantic web can be eased by the use of learning techniques. To completely induce a probabilistic ontology (that is, an ontology encoded through a probabilistic description logic) from data, two basic tasks must be solved: (1) learning concept definitions and (2) learning probabilistic inclusions. In this paper we propose and test an algorithm that learns concept definitions using an inductive logic programming approach and then learns probabilistic inclusions using relational data.
|Original language||American English|
|Number of pages||12|
|State||Published - 1 Dec 2010|
|Event||CEUR Workshop Proceedings - |
Duration: 1 Jan 2016 → …
|Conference||CEUR Workshop Proceedings|
|Period||1/01/16 → …|