The RETE algorithm has been used to implement first-order logic based inference engines and its object-oriented extension allows to reason directly over entities rather than predicates. One of the limitations of FOL is its inability to deal with uncertainty, although it exists in many forms and it is typical of the way humans reason. In this paper, the steps of a general uncertain reasoning are outlined, without choosing a specific type or representation of uncertainty. Then, the process is translated into a further extension of the RETE networks, showing a possible architecture allowing a Rete-OO based engine to reason with uncertain rules. This architecture is being implemented in the Drools rule engine.
Adding Uncertainty to a Rete-OO Inference Engine / D.Sottara; P. Mello; M. Proctor. - STAMPA. - 5321:(2008), pp. 104-118. (Intervento presentato al convegno INTERNATIONAL SYMPOSIUM, RULEML 2008 tenutosi a Orlando FL USA nel 30-31 October 2008) [10.1007/978-3-540-88808-6-13].
Adding Uncertainty to a Rete-OO Inference Engine.
SOTTARA, DAVIDE;MELLO, PAOLA;
2008
Abstract
The RETE algorithm has been used to implement first-order logic based inference engines and its object-oriented extension allows to reason directly over entities rather than predicates. One of the limitations of FOL is its inability to deal with uncertainty, although it exists in many forms and it is typical of the way humans reason. In this paper, the steps of a general uncertain reasoning are outlined, without choosing a specific type or representation of uncertainty. Then, the process is translated into a further extension of the RETE networks, showing a possible architecture allowing a Rete-OO based engine to reason with uncertain rules. This architecture is being implemented in the Drools rule engine.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
