Matching points of interest with user context: an ANN approach

In this paper, the design and development of an artificial neural network (ANN) for similarity value calculation in a context-aware system is proposed. This neural network is used by the neural agent of the iConAwa system. Since iConAwa is an intelligent, context-aware, multiagent system, it provides mobile users with context-aware information and services, and also provides communication with each other. Context and points of interest are modeled in a flexible and extensible way by using ontologies. iConAwa derives high-level implicit context from low-level explicit context by inference performed over the context ontology. This approach decouples context reasoning from the source code of the system. With the addition of a neural agent, which uses an ANN, the system has learning capability. By using a neural network for similarity value calculation, the system can adapt to the needs of different people. System owners can introduce their own similarity metric considering their own requirements, which further improves the iConAwa system. Thus, the extended iConAwa system combines expert system characteristics with the capability to learn.In this paper, the design and development of an artificial neural network (ANN) for similarity value calculation in a context-aware system is proposed. This neural network is used by the neural agent of the iConAwa system. Since iConAwa is an intelligent, context-aware, multiagent system, it provides mobile users with context-aware information and services, and also provides communication with each other. Context and points of interest are modeled in a flexible and extensible way by using ontologies. iConAwa derives high-level implicit context from low-level explicit context by inference performed over the context ontology. This approach decouples context reasoning from the source code of the system. With the addition of a neural agent, which uses an ANN, the system has learning capability. By using a neural network for similarity value calculation, the system can adapt to the needs of different people. System owners can introduce their own similarity metric considering their own requirements, which further improves the iConAwa system. Thus, the extended iConAwa system combines expert system characteristics with the capability to learn