On Intentional and Social Agents with graded Attitudes

ANA CASALI

Serie de Monografías del Instituto de Investigación en Inteligencia Artificial, CSIC

Lugar: Barcelona; Año: 2009 p. 260

978-84-00-08850-7

The central contribution of this dissertation is the proposal of a graded BDI agent model(g-BDI), specifying an architecture capable of representing and reasoning with graded mentalattitudes. We consider that making the BDI architecture more flexible will allow us todesign and develop agents capable of improved performance in uncertain and dynamic environments, serving other agents (human or not) that may have a set of graded motivations.In the g-BDI model, the agent graded attitudes have an explicit and suitable representation.Belief degrees represent the extent to which the agent believes a formula to betrue. Degrees of positive or negative desires allow the agent to set dierent levels of preferenceor rejection respectively. Intention degrees also give a preference measure but, inthis case, modelling the cost/benet trade o of achieving an agent´s goal. Then, agentshaving dierent kinds of behaviour can be modelled on the basis of the representation andinteraction of their graded attitudes. The formalization of the g-BDI agent model is basedon Multi-context systems and in order to represent and reason about the beliefs, desiresand intentions, we followed a many-valued modal approach. Also, a sound and completeaxiomatics for representing each graded attitude is proposed. Besides, in order to cope withthe operational semantics aspects of the g-BDI agent model, we rst dened a Multi-contextcalculus for Multi-context systems execution and then, using this calculus we give this agentmodel computational meaning.Furthermore, a software engineering process to develop graded BDI agents in a multiagentscenario is presented. The aim of the proposed methodology is to guide the design ofa multiagent system starting from a real world problem. Through the development of aTourism recommender system, where one of its principal agents is modelled as a g-BDIagent, we show that the model is useful to design and implement concrete agents.Finally, using the case study we have made some experiments concerning the exibility and performance of the g-BDI agent model, demonstrating that this agent model is useful to develop agents showing varied and rich behaviours. We also show that the results obtainedby these particular recommender agents using graded attitudes improve those achieved byagents using non-graded attitudes.