Yuu Nakajima

Learning From Humans: Agent Modeling With Individual Human Behaviors

Multiagent-based simulation (MABS) is a very active interdisciplinary area bridging multiagent research and social science. The key technology to conduct truly useful MABS is agent modeling for reproducing realistic behaviors. In order to make agent models realistic, it seems natural to learn from human behavior in the real world. The challenge presented in this paper is to obtain an individual behavior model by using participatory modeling in the traffic domain. We show a methodology that can elicit prior knowledge for explaining human driving behavior in specific environments, and then construct a driving behavior model based on the set of prior knowledge. In the real world, human drivers often perform unintentional actions, and occasionally, they have no logical reason for their actions. In these cases, we cannot rely on prior knowledge to explain them. We are forced to construct a behavior model with an insufficient amount of knowledge to reproduce the driving behavior. To construct such individual driving behavior model, we take the approach of using knowledge from others to complement the lack of knowledge from the target. To clarify that the behavior model including prior knowledge from others offers individuality in driving behavior, we experimentally confirm that the driving behaviors reproduced by the hybrid model correlate reasonably well with human behavior.

Disaster Evacuation Guide: Using a Massively Multiagent Server and GPS Mobile Phones

The ubiquitous environment enables us to build systems that provide individual users with personalized navigation services in cities. In developing such a system, it is necessary to estimate the influence and the movement of its users and to get feedback. However, it is difficult to perform tests on such a system given the large number of human subjects involved and its scale, which matches that of a major city. One possible solution is proposed herein, the augmented experiment; it combines a multiagent simulation with a small-scale experiment performed with human subjects. In the experiment, the movements of agents that simulate users are shown to human subjects in order to give them the impression that the environment is populated with a large number of users. In this research, we build an evacuation guide system based on GPS-capable cellular phones and perform an augmented experiment wherein human subjects and evacuee agents are directed while the status of the simulation is passed to the human subjects. Interviews of the human subjects confirmed that the augmented experiment successfully gave the impression that a large number of users were present

Massive Multiagent-Based Urban Traffic Simulation with Fine-Grained Behavior Models

As it is getting easier to obtain reams of data on human behavior via ubiquitous devices, it is becoming obvious that we must work on two conflicting research directions for realizing multiagent-based social simulations; creating large-scale simulations and elaborating fine-scale human behavior models. The challenge in this paper is to achieve massively urban traffic simulations with fine-grained levels of driving behavior. Toward our objective, we show the design and implementation of a multiagent-based simulation platform, that enables us to execute massive but sophisticated multiagent traffic simulations. We show the capability of the developed platform to reproduce the urban traffic with a social experiment scenario. We investigate its potential to analyze the traffic from both macroscopic and microscopic viewpoints.

Caribbean/Q: A Massively Multi-Agent Platform with Scenario Description Language

Making a truly useful massively multi-agent system is difficult since the actions of the full ensemble of agents cannot be controlled by designing just one agent. It is critical to control all the agents by using protocols that describe the interaction of agents and the environment in a top-down approach. We introduce a system that uses interaction protocol descriptions and has the capability of controlling hundreds of thousands of agents. This makes it feasible to realize a mega-scale navigation system that can assist the inhabitants of a small city. In developing a massively multiagent system, protocol design and agent development need to be separated to allow specialists to work in concert with one another while honing the different technologies. As a platform for mega-scale navigation system, we devise an architecture for multiagent platforms where the execution of agents scenario and the implementation of agents are explicitly separated. This paper also gives the evaluation and an application example using the platform.

Multi-model based simulation platform for urban traffic simulation

Multiagent-based simulations are regarded as a useful technology for analyzing complex social systems; for example, traffic in a city. Traffic in a city has various aspects such as route planning on the road network and driving operations on a certain road. Both types of human behavior are being studied separately by specialists in their respective domains. We believe that traffic simulation platforms should integrate the various paradigms underlying agent decision making and the target environment. We focus on urban traffic as the target problem and attempt to realize a multiagent simulation platform based on the multi-model approach. While traffic flow simulations using simple agents are popular in the traffic domain, it has been recognized that driving behavior simulations with sophisticated agents are also beneficial. However, there is no software platform that can integrate traffic simulators dealing with different aspects of urban traffic. In this paper, we propose a traffic simulation platform that can execute citywide traffic simulations that take account of the aspects of route selection on a road network and driving behavior on individual roads. The proposed simulation platform enables the multiple aspects of city traffic to be reproduced while still retaining scalability.

A Driver Modeling Methodology Using Hypothetical Reasoning for Multiagent Traffic Simulation

We propose how to acquire drivers individual operation models using the three-dimensional driving simulator in order to implement distinct personalities on each agent. In this paper, operation models are defined as sets of prioritized operation rules, each of which consists of the world as observed by a driver and his/her next operation according to the observation. Each driver might have different set of rules and their priorities. We apply a method to acquire individual operation models using hypothetical reasoning. Because of the method, we are able to obtain operation models which can explain drivers operation during driving simulation. We show some operation models acquired from aged/young human drivers, and then clarify the proposed method can catch each drivers characteristics.

Multiagent Gaming System for Multilingual Communication

The internationalization of economic and social activities is forcing people who use different languages and belong to different cultures to collaborate far more often. An analysis of the status of multilingual communications among such people is required. We apply gaming simulation methods to analyze this topic. In this study, we design an environment that can game multilingual communication online by describing just simple game scripts. This advance is significant as people with domain knowledge of the applied problem are not always computer experts.

Attempt to mitigate congestion by providing latent information with walk-rally application

At events and sightseeing spots, congestion can be a factor that lowers customer satisfaction, so there is need for methods that mitigate congestion. Currently, there are methods for mitigating congestion such as changing the layout of an event and restricting customer behavior. However, there are problems in that it is difficult to reduce congestion at events in which the layout cannot be changed, and there is a risk the satisfaction level will be reduced due to restricting customer behavior. Therefore, we propose a method aimed at reducing congestion by allowing users to voluntarily select a detour to avoid congestion by providing users with not only obvious information but latent information with a walk-rally application. In demonstration experiments at real events and sightseeing spots, we confirmed that the proposed method is effective at reducing congestion and clarified how to use latent information effectively for mitigating congestion.

Gaming for Language Services

Service-oriented computing environments (SoCEs) such as the Language Grid can be regarded as the synthesis of their individual services. However, to make sustainable SoCEs, the user will want assurance that the billing structure is valid of and apportionment of cost burden among users is fair. To analyze the factors involved, we conduct donation and investment games for a machine translation service, where the service users participate in gaming. The results confirm the existence of users that actively try to make donations and social investments, users that remain passive to this kind of service, as well as users who tend to free ride on other users. Furthermore, we find that setting bonuses based on the total amount of donation and investment is effective in incentivizing some players to donate and invest. We also show how to reduce the cost of developing and executing gaming exercises for domain practitioners or experts. We define a game as a workflow of collaborative tasks executed by players. We develop game definition criteria to simplify game descriptions. We develop a gaming environment that enables web-based games to be implemented by using the game definitions.

Field-Oriented Service Design: A Multiagent Approach

Service has been considered as value co-creation through the cooperation of service providers and customers. This paper, however, focuses on service design in problem fields where complex issues exist among various stakeholders, where identifying service providers and customers is not a simple process. In other words, we focus on a very early stage of service design with huge ambiguities; we call it field-oriented service design. A typical case is introducing new services in developing countries. The main issue here is to create new services compatible with existing services through action research that considers a wide variety of regional, national and global stakeholders. It is often difficult to identify the influence of/to the services to be designed due to the differences in culture, language and business customs. As a result, unexpected interdependencies among services together with stakeholders are often revealed during the process of action research. To resolve this ambiguity in the design process, we propose a multiagent approach that couples role playing games with participatory simulations; it is based on our experiences in agricultural support projects in Southeast Asia.