TY - GEN
T1 - A framework for validating task assignment in multiagent systems using requirements importance
AU - Nakagawa, Hiroyuki
AU - Yoshioka, Nobukazu
AU - Ohsuga, Akihiko
AU - Honiden, Shinichi
PY - 2012/12/1
Y1 - 2012/12/1
N2 - Multi-agent systems (MASs) are one of the effective approaches for dealing with the recent increase in software complexity and their autonomy. In the MAS research community, there has recently been increasing interest in the adoption of requirements engineering techniques to bridge the gap between the system requirements and the system design. One of the most important tasks based on the requirements description in the MAS design activity is the extraction of roles, which are the fundamental components of multi-agent systems, from it. It is also important to comprehend the relative degree of responsibility of the individual roles. The comprehension helps the developer decide the system architecture and discuss the performance and stability of the system. We introduce the concept of importance as a quantitative metric and an evaluation framework for the extraction of a suitable role set for the system and the task assignment to these roles. The importance is propagated from the requirements to the roles through their assigned tasks. We demonstrate the effectiveness of our framework through a case study and show that our metric and evaluation framework help not only to identify the importance of each role, but also to determine the system architecture.
AB - Multi-agent systems (MASs) are one of the effective approaches for dealing with the recent increase in software complexity and their autonomy. In the MAS research community, there has recently been increasing interest in the adoption of requirements engineering techniques to bridge the gap between the system requirements and the system design. One of the most important tasks based on the requirements description in the MAS design activity is the extraction of roles, which are the fundamental components of multi-agent systems, from it. It is also important to comprehend the relative degree of responsibility of the individual roles. The comprehension helps the developer decide the system architecture and discuss the performance and stability of the system. We introduce the concept of importance as a quantitative metric and an evaluation framework for the extraction of a suitable role set for the system and the task assignment to these roles. The importance is propagated from the requirements to the roles through their assigned tasks. We demonstrate the effectiveness of our framework through a case study and show that our metric and evaluation framework help not only to identify the importance of each role, but also to determine the system architecture.
UR - http://www.scopus.com/inward/record.url?scp=84887245918&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84887245918&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-25920-3_32
DO - 10.1007/978-3-642-25920-3_32
M3 - Conference contribution
AN - SCOPUS:84887245918
SN - 9783642259197
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 443
EP - 458
BT - Principles and Practice of Multi-Agent Systems - 13th International Conference, PRIMA 2010, Revised Selected Papers
T2 - 13th International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2010
Y2 - 12 November 2010 through 15 November 2010
ER -