Interoperation in Complex Information Ecosystems

Summary

Individuals, enterprises and policy makers increasingly rely on data to stay informed and make decisions. The amount of available digital data grows at a tremendous pace. At the same time, the number of systems providing and processing data increases, leading to complex information ecosystems with large amounts of data, a multitude of stakeholders, and a plethora of data sources and systems. Thus, there is an increasing need for integration of information and interoperation between systems.

Due to the ubiquitous need for integration and interoperation, many research communities have tackled the problem. Recent developments have established a pay-as-you-go integration model, where integration is seen as a process starting out with enabling only basic query functionality over data and iteratively spending targeted integration effort as the need for more complex queries arises. Such an ad-hoc model is in contrast to previous integration models which required the construction of a mediated schema and the integration of schema and data before any queries -- even simple ones -- could be answered. The move towards less rigid integration systems can be traced back to many communities: the database community established Dataspaces as a new abstraction for information integration; the Semantic Web community provided ontologies and logic-based modelling in a web context; finally, the Web community established the Hypermedia principle which enables decentralized discovery and ad-hoc unidirectional interlinking in very large information systems.

Current systems for data integration focus on query-related aspects. However, to enable real interoperation, updates and invocation of functionality are required. Mobile applications, for example, require both access to information and functionality. We want to broaden the scope of research on data integration towards a vision of interoperation between systems (i.e., systems that no only exchange and integrate their data but also link functionality) and investigate how an iterative model can be established for the interoperation of systems.

The seminar has multiple objectives:

• to bring together researchers from these diverse communities to identify common themes and to exploit synergies,
• to develop the theoretical foundations and an understanding of architectures and methods,
• to develop a research agenda and road-map towards a vision of web-scale integration and interoperation.

The seminar has been supported by ONR Global, grant N62909-13-1-C161.