@Article{pakinsight,
AUTHOR = {},
TITLE = {Hospital Site Selection Using a BDI Agent Model},
JOURNAL = {International Journal of Geography and Geology},
VOLUME = {2},
YEAR = {2013},
NUMBER = {4},
PAGES = {36-51},
URL = {http://www.pakinsight.com/archive/10/04-2013/4},
ISSN = {2305-7041},
ABSTRACT = {This paper presents a newly developed Belief-Desire-Intention (BDI) Agent-based model for estimating suitable hospital sites. Our model makes use of existing geospatial functions and a novel BDI architecture of agent techniques. More specifically, the fundamental concepts of practical reasoning architecture such as belief, desire, intention, along with commitment, and interaction have been combined with analyses and applications of Geographical Information System (GIS). The proposed model can be customized for a wide range of decision making problems in GIS, one of which is site selection. In this model, minimal travel time, air pollution and land cost are considered as the goals of agents, and then the agents observe, and believe in the environment. The agent then determines the intention to implement on the environment for achieving their desires. The desires are generated from agents’ goals. The interactions among agents are considered as a part of process for achieving contemporarily goals. In this paper, the fundamental components of agent such as observation, belief, desire, intention, commitment, and interaction are introduced spatially, and a BDI-GIS model is defined based on these components. The Desktop GIAgent software, introduced in this paper, has the advantage of using agents for spatial analysis. The interface helps guiding decision makers through the sequential steps for site selection, namely; importing data, defining goals, determining actions and identifying the agent’s characteristics. For demonstrating the robustness of our new model, a case study was planned and executed in Tehran, Iran. The efficiency of the BDI-GIS model in the decision making process for selecting suitable hospital sites was also demonstrated based on the characteristics of the agents and the types of their interactions.},
DOI = {10.18488/journal.10/2013.2.4/10.4.36.51}
}