Business and model-driven development of BDI multi-agent systems

被引:31
作者
Wautelet, Yves [1 ]
Kolp, Manuel [2 ]
机构
[1] Katholieke Univ Leuven, Louvain, Belgium
[2] Catholic Univ Louvain, Louvain, Belgium
关键词
Service modeling; BDI software system; I*; Actor responsibility assignment; Multi-agent system; Model-driven engineering; Agent-based development; TROPOS; METAMODEL;
D O I
10.1016/j.neucom.2015.12.022
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Model-driven development allows IT profe.ssionals to specify the system functionality, organization and behavior in a logical or platform-independent manner. Modeling using services allows domain analysts to focus on the added-value and core business the enterprise offers to its stakeholders. Those services are coarse-grained elements able to encapsulate a composition of business process models. The framework presented in this paper provides models together at strategic, tactical and operational levels to develop an agent-oriented software system. The strategic level is concerned with long-term decisions; this top-level uses a service model to understand the business' high-level (added) values as well as the Quality Expectations and the threats they face. The tactical level is concerned with a broader description of the business processes automated by the system; the i* strategic dependency and rationale models are used here to further document the service behavior. Actors' accountability and responsibility can be determined in the visual representation of these strategic and tactical levels. Finally, i* models are mapped into a set of operational models to document the (multi-agent) system behavior when achieving modeled functionalities. These operational models instantiate the Belief/Desire/Intentions (BDI) paradigm proposing entities - the agents - mapping as closely as possible the real life organization. The paper thus builds a business-driven transformation process leading to a run-time agent-architecture in a single and common framework. It both uses existing models and introduces or refines existing ones to dispose of a method ensuring better alignment and traceability between the business and the if system. (C) 2015 Elsevier B.V. All rights reserved.
引用
收藏
页码:304 / 321
页数:18
相关论文
共 45 条
  • [1] MASD: Multi-agent systems development methodology
    Abdelaziz, T.
    Elammari, M.
    Unland, R.
    Branki, C.
    [J]. MULTIAGENT AND GRID SYSTEMS, 2010, 6 (01) : 71 - 101
  • [2] [Anonymous], 2001, 5 INT C AUT AG
  • [3] [Anonymous], 2012, ENTERPRISE SYSTEMS M
  • [4] [Anonymous], 2009, An introduction to multiagent systems
  • [5] [Anonymous], 2011, Social modeling for requirements engineering
  • [6] [Anonymous], 2009, MAN SUCC PROJ PRINCE
  • [7] [Anonymous], 2007, DEV MULTIAGENT SYSTE
  • [8] A model driven engineering process of platform neutral agents for ambient intelligence devices
    Ayala, Inmaculada
    Amor, Mercedes
    Fuentes, Lidia
    [J]. AUTONOMOUS AGENTS AND MULTI-AGENT SYSTEMS, 2014, 28 (02) : 214 - 255
  • [9] Agent UML:: A formalism for specifying multiagent software systems
    Bauer, B
    Müller, JP
    Odell, J
    [J]. INTERNATIONAL JOURNAL OF SOFTWARE ENGINEERING AND KNOWLEDGE ENGINEERING, 2001, 11 (03) : 207 - 230
  • [10] Bratman M., 1999, INTENTION PLANS PRAC