Logics and Networks for Human Reasoning

We propose to model human reasoning tasks using completed logic programs interpreted under the three-valued Lukasiewicz semantics. Given an appropriate immediate consequence operator, completed logic programs admit a least model, which can be computed by iterating the consequence operator. Reasoning is then performed with respect to the least model. The approach is realized in a connectionist setting. Key words: Human Reasoning, Logic Programs, Connectionist Models

• Type: Conference
• Year: 2009
• Where: ICANN
• Authors: Steffen Hölldobler, Carroline Dewi Puspa Kencana Ramli
• Keywords: Artificial Intelligence, Consequence Operator, Human Reasoning, ICANN 2009, Logic Programs