登入選單
返回Google圖書搜尋
出版INTECH Open Access Publisher, 2008
ISBN95376190369789537619039
URLhttp://books.google.com.hk/books?id=tC2yoAEACAAJ&hl=&source=gbs_api
註釋In this chapter, the application of multi-agent systems to tackle the software engineering task was outlined. The concentration was on the employment of agent technology in order to deal with distributed software systems and mainly distributed database applications and web applications. The rationale behind utilizing agent technology has to do with the multi-tier architecture and the associated inherent complication of distributed applications and the required interoperability of software resources belonging to potentially disparate application components and disparate domains. To meet these requirements, agents offer a unified platform of interaction through agent communication. The current research status can be classified according to two principal tracks. The first one has as a goal to provide an agent infrastructure to support software testing. This is realised by suggesting multi-agent frameworks that can be used as a model to build agent systems for testing service-oriented web applications. The second one has a more applied nature. This research track aims at presenting an agent system for tackling the issues of software maintenance and testing of distributed applications. Analysing the aforementioned research attempts some general comments can be stated. A first and important comment is that all approaches have a quite narrow scope. On the one hand, the application domain is related to web services, web applications and database applications. The only exception is the work of (Yamany et al., 2006) but even in this case only 3-tier applications are considered. These domains have a surely specific nature even though they provide a solid basis for introducing the existing attempts. Moreover, this restriction is made clearer by the fact that the software engineering process is not covered in its complete form. Almost all attempts target software testing with the exception of (Gardikiotis et al., 2007a) where software maintenance is also treated in depth. The above work is also the only one where the existing platform has proven its extensibility by including generic software engineering agents. Focusing on infrastructural approaches, the work of both (Ma et al., 2007) and (Xu et al., 2006) has a very specific objective which is to support agent collaboration. Besides this commonality, the research of (Xu et al., 2006) is more tailored to software testing encompassing the notion of test tasks while the one of (Ma et al., 2007) recommends an.