TUWIEN DReW

From ONTORULE Show Case

Motivation

DReW is:

1) a prototype DL reasoner over LDL+ ontologies: taking as input a conjunctive query and an ontology in OWL 2 syntax extended for the complex role expressions of LDL+, DReW checks whether the ontology is in the LDL+ fragment. If it is, it is translated according to the format suitable for the specified Datalog reasoner (DLV in our case). The reasoner returns in this case the answers to the conjunctive query.

2) a prototype reasoner for dl-programs over LDL+ ontologies under well-founded semantics: taking as input a dl-program over an ontology in OWL 2 syntax extended for the complex role expressions of LDL+, DReW checks whether the ontology is in the LDL+ fragment. If it is, it is translated according to the format suitable for the specified Datalog reasoner (DLV in our case) and the well-founded model of the resulted program is computed.

Description

DL-programs are a loose coupling of OWL ontology and Answer Set Programming. The traditional ways of reasoning over dl-programs use a native DL reasoner to reason about the ontology part, and use a native rule reasoner to deal with the rule part. Due to the interaction between the two parts, many calls to the DL and rule reasoner generally can not be avoided, which causes overhead. To remedy this, we propose a new strategy, called inline evaluation, which rewrites the whole dl-program KB into a new KB using only a rule formalism. Hence we can use a single rule reasoner to do the reasoning tasks, and improve the efficiency of reasoning.

The underlying theoretical work is reported in Deliverable D3.3.

Usage

1. Download DReW from http://www.kr.tuwien.ac.at/research/systems/drew/download.html

2. You need to download the latest dlv which supports well-founded semantics. Download the dlv file, rename it to dlv_wf, and add the dictory of dlv_wf to $PATH.

3. In the distribution you can find examples in the drew-examples, drew-ldlp-dl-reasoner-evaluation, and drew-ldlp-dl-reasoner-evaluation modules.