Using architecture integration patterns to compose enterprise mashups

Enterprise mashups deal with corporate data and various sources of information to compose new value-added applications. The architecture design of enterprise mashups encompasses integration issues-it needs to integrate heterogeneous data and/or compose new situational applications from existing infrastructure. We envisage that architecture integration patterns can be applied not only as architecture solutions to mashup development, but also to help develop practical mashup techniques. In this paper, we combine several common architecture integration patterns, namely pipes and filters, data federation, and model-view-control to compose enterprise mashups. A number of techniques are also developed to customize these patterns for specific mashup needs. We illustrate our approach with a property valuation service derived from a real-world setting.