Thomas Farrenkopf

AGADE Using Personal Preferences and World Knowledge to Model Agent Behaviour

BDI agents provide a common well established approach for building multi-agent simulations. In this paper we demonstrate how semantic technologies can be used to model agent behaviour. Beliefs, desires and intentions are mapped flexibly to corresponding OWL ontologies structured in layers. This reduces JAVA coding efforts significantly. Reasoning mechanisms and rule evaluation are used to compute agent behaviour by deriving an agent’s actions from declaratively formulated rules. An agent’s knowledge of its environment and its personal preferences can be expressed and human behaviour can be simulated. The approach is implemented in an integrated tool for running round based agent simulations (AGADE).

Ontology Based Business Simulations

Within business games there is a need to provide realistic feedback for decisions made, if such business games are to continue to remain relevant in increasingly complex business environments. We address this problem by using software agents to simulate individuals and to model their actions in response to business decisions. In our initial studies we have used software agents to simulate consumers who make buying decisions based on their private preferences and those prevalent within their social network. This approach can be applied to search for behavioural patterns in social structures and to verify predicted values based on a priori theoretical considerations. Individual behaviour can be modelled for each agent and its effects within the marketplace can be examined by running simulations. Our simulations are founded upon the BDI software model (belief-desire-intention) combined with ontologies to make world knowledge available to the agents which can then determine their actions in accordance with this knowledge. We demonstrate how ontologies can be integrated into the BDI concept utilising the Jadex agent framework. Our examples are based upon the simulation of market mechanisms within the context of different industries. We use a framework, developed previously, known as AGADE within which each agent evolves its knowledge using an ontology maintained during the simulation. This generic approach allows the simulation of various consumer scenarios which can be modelled by creating appropriate ontologies.

Towards Reducing Complexity of Multi-agent Simulations by Applying Model-Driven Techniques

Creating multi-agent simulations is a challenging task often requiring programming skills at the professional software developer level. Model driven methods of software development are an appropriate tool for reducing the complexity of the development process of such simulations. The modeller is relieved from implementing time consuming programming details and can concentrate on the application itself. We present the domain specific language Athos with which network based traffic simulations can be created declaratively. The models are platform independent and executable code can be generated for two popular multi-agent platforms. We use a simple yet illustrative example to show how Athos can be applied.

ATHOS - A Domain-Specific Language for Multi-agent Simulations

Creating multi-agent simulations is a challenging task often requiring programming skills at the professional developer level which domain experts scarcely possess. We present a model-driven approach that relieves agent experts from time-consuming, error-prone implementation tasks and allows them to focus on the application itself. With our domain specific language Athos, network-based traffic simulations can be created declaratively. The models are platform independent and executable code can be generated for two popular multi-agent platforms.

A Domain-Specific Language For Routing Problems.

Vehicle Routing Problems (VRPs) are commonly used as benchmark optimisation problems and they also have many applications in industry. Using agent-based approaches to solve VRPs allows the analysis of dynamic VRP instances that incorporate congestion effects. By using a domain-specific language as part of a model-driven approach, routing problems can be modelled in an abstract form that does not contain implementation and other technical details. With such a tool domain experts can concentrate on the actual modelling task without being distracted by low-level intricacies. We present the DSL Athos in which computational and platform independent routing problems can be defined. The DSL offers an efficient way to model problems with seamless integration of established optimisation methods. Generators create executable code for several agent based platforms. Proof of concept is given by applying the tools to the Oliver 30 TSP and an instance of a dynamic TSP.

Demo Paper: AGADE

Simulations of real world scenarios often require considerably large numbers of agents. With increasing level of detail and resolution in the underlying models machine limitations both in the aspect of memory and computing power are reached. Even more when additional features like reasoning mechanisms of semantic technologies are used as in the AGADE framework where we have extended the principal BDI paradigm with an interface to OWL ontologies. We have observed that the extensive use of ontologies results in high memory consumption due to the large number of String objects used in the reasoning process and caching mechanisms of the OWL API. We address this issue by running simulations in a highly distributed environment. In this paper we demonstrate how we enabled AGADE to be run in such an environment and the necessary architectural modifications. Furthermore, we discuss the potential size of simulations that can be run in such a setting.