CIMSim 2012 : 4th international conference on computational intelligence, modelling and simulation : proceedings

International audience; Studies in predicting rice pest outbreaks use regression models to explore relationships between rice pest population density and environmental factors. Agent-based models provide a flexible assessment framework to help stakeholders in planning rice pest management strategies. This paper proposes a general method to determine rice pest risk index and design a structure of agents that have an ability of aggregating information from other agents in agent-based models, OWA-based method is experimented on multi-criteria decision making agents to determine rice pest risk index of each rice area. A case study is carried out on risk assessment of brown plant hopper in the Mekong Delta region, simulation results in this study show that rice pest risk index aggregated from criteria and agent-based approach are useful tools to support stakeholders in rice pest management.