There have...

It is important to understand how people act in urban areas in order to design any social and economic systems. In this paper we focus on the spatial phase of peoples act, in other word, how they move around. The need to comprehend the way in which people move through urban areas leads to the desire to predict their movement, such as an assessment of impacts of new policy or an identification of the optimum location for a new shop. This paper describes a structure of agent-based behavior model, which is intended as the rough framework of such predictive models. The model suggested here is based on some analysis of survey data sets, including real trajectories, mental process and viewing fields of pedestrians. It is because of the assumption that pedestrian movement should be influenced by many factors such as configuration and the location of attractions, previous knowledge of the place, quality and quantity of the information they acquire and so on. So as the first step of making a multi-agent based model, some ideas of current human behavioral models in other different fields such as social psychology and cognitive sciences were reviewed and introduced to decide what kind of factors should be included in the model to cover most of the possible influence on pedestrian movement. In addition to these, demographic profiles of the pedestrians, the attributes or preferences of guests, are included in our model framework. We will develop a series of Multi-agent simulation systems and their applications, some of which are currently still under development but will be finalized in near future to examine these possible factors. The results from these simulations will show which factor(s) really have major influence on spatial movement and will help us with settin