22nd International Conference on Principles and Practice of Multi-Agent Systems (PRIMA 2019)
2019
Explicitly modelling the environment as a first-class abstraction is essential in the design of situated multi-agent systems (MAS): according to the A&A meta-model, the notion of artefact can be used for that purpose, as the abstraction in charge of augmenting agents’ capabilities and relieving their cognitive workload. Following Clark and Chalmers’ idea of “active externalism”, we argue that artefacts can participate in agents’ cognitive processes by acting as delegates for agents’ intelligent behaviour—in particular when they are logic-based artefacts. Along this line, in this paper we define micro-intelligence as the externalised rationality complementing agents’ cognition, and propose situated logic programming as the conceptual framework for delivering micro-intelligence in MAS, LPaaS as the reference architecture, and LPaaS-REST as its technological embodiment.
keywords
distributed cognition; micro-intelligence; externalised rationality in A&A; situated logic programming; LPaaS; LPaaS-REST
origin event