In this paper we present the linked data sets produced by the eagle-i project. We describe the content, the features and some of the applications currently leveraging these datasets.

Core facilities need customers, yet an average researcher is aware of a very small fraction of the facilities available near them. Diligent facilities combat this knowledge gap by broadcasting information about their facility either locally, or by publishing information on one or multiple public-facing websites or third-party repositories of information (e.g. VGN cores database or Science Exchange). Each additional site of information about a facility increases visibility but also impairs their ability to maintain up-to-date information on all sites. Additionally, most third-party repositories house their data in traditional relational databases that are not indexable by common search engines. To start addressing these problems, the eagle-i project (an free, open-source, open-access, publication platform), has begun integrating its core facility database with external websites and web applications allowing them to synchronize their information in real-time. We present here two experimental integrations. The Harvard Catalyst Cores webpage originally required independent updates which were not within the direct control of the core directors themselves. The eagle-i linked open data architecture developed now allows the Catalyst cores page to pull information from the Harvard eagle-i server and update all data on it's page accordingly. Additionally, Harvard's “Profiles” web application references eagle-i data and links resource information from eagle-i to personnel information in the Profiles database. Because of these direct links, updating information in Harvard's eagle-i server (which can be accessed directly by facility directors through an account on the eagle-i SWEET), automatically updates information on the Catalyst Cores webpage and updates resource counts linked to a researcher's public profile. This functionality of the eagle-i platform as a central “point-of-truth” for information has the potential to drastically reduce the effort required to efficiently disseminate core facility information.

The eagle-i Consortium ("http://www.eagle-i.org/home":www.eagle-i.org/home) comprises nine geographically and ethnically diverse universities across America working to build a federated network of research resources. Biomedical research generates many resources that are rarely shared or published, including: reagents, protocols, instruments, expertise, organisms, training opportunities, software, human studies, and biological specimens. The goal of eagle-i is to improve biomedical research by helping researchers more easily find scientific resources that are difficult to discover, reducing time-consuming and expensive duplication of resources. Now in early development, the system will ultimately expand to include research resources at other universities following the end of the two-year pilot phase. An application ontology is being developed to enable representation of core facility and research lab resources in the eagle-i repository, leading to more effective searches and better linkage between data types. The eagle-i ontology will guide users to valid queries via auto-suggestion, ontology browsing, concept-based search, and synonym expansion. The ontology development effort is being guided by active discussions within the ontology community and brings together relevant preexisting ontologies in a logical framework. Components of the data entry and search interfaces are generated directly from the ontology, which allows rapid change in response to user needs and ontology evolution. Each eagle-i institution will populate and manage a local repository using data collection and curation tools. To enhance the quantity and quality of data, the data tools will take advantage of the ontology to support semi-automated annotation of resources. NIH/NCRR ARRA award U24RR029825.

We propose a novel approach to the problem of expertise mining in an enterprise, taking advantage of online social applications deployed within the enterprise. Based on the assumption that the users’ interactions with such social software reflect to some extent their expertise, we devise a probabilistic method for identifying the main areas of expertise of users based solely on their set of tags extracted from a social bookmarking system. We base our approach on statistical language models, which we adapt to fit our unique setting. We train and validate our model on a real world dataset extracted from two IBM-internal applications. Our results show that our approach provides a viable alternative to other methods that rely on documents extracted from the enterprise corpora.

We propose Commune, a workflow management system that supports a mobile workforce. We introduce the "mini-workflow," a network-isolated subcomponent of a business workflow process. Miniworkflows enable subsets of business workflow to be off-loaded into mobile client environments by leveraging web services. We describe how miniworkflows overcome the challenges of intermittent connectivity and we describe our BPEL-based implementation.

End-system resource management is a key issue for obtaining satisfying QoS in multimedia applications. As these applications are being incorporated into workstations, problems regarding efficient and well suited operating system mechanisms to cope with multimedia data constraints have been outlined. In this paper we present a runtime kernel analysis of an adaptive multimedia application dealing with compressed audio and video, which is structured as different communicating processes, running on Solaris 2.4. These processes are not independent, and their interactions determine the application's overall performance. We show that even if application structure is optimised, the fact that synchronisation between processes exists but is not considered by the process scheduling policy, results in unnecessary data copying and image rejection and consequently poor performances. Our measurements of application performance highlight the need for a system level synchronisation support.

The fluid computing middleware allows application data to flow, as a liquid, between devices. The middleware, based on the model-view-controller design pattern, transforms a collection of devices into a cooperative platform and provides application programmers with appropriate abstractions to use generic data management services, such as data replication and synchronization of arbitrary data structures. Relieved from managing this underlying complexity, programmers can focus on developing new application functionality. In particular, they can program applications that seamlessly transition between devices and connectivity degrees, by relying on the fluid real-time optimistic replication mechanism. By shifting the focus from the network and devices to the data, fluid computing enables new interaction paradigms, where users can access and manipulate their data from any of their devices transparently. Applications, then, become multidevice - a characteristic that can be applied to both single-user and collaboration scenarios.

This paper proposes a novel application of the popular content syndication format RSS (really simple syndication) to improve the user experience of humans interacting with electronic workflows. We will discuss this in relation to a more general solution, formidable, for developing and deploying electronic business processes that consist of a mixture of automatic and human-facing steps

In the typical Web application, a client renders markup and the application or service is implemented as a set of tiered functions in the network. However, clients can contain resources useful for an application and network connections cannot always be assumed. In this paper, we consider extending the reach of a Web application to include: 1) access to and use of local client resources, and 2) operation while disconnected from the network. We, however, try to preserve desirable programming model and management characteristics of Web applications. We propose a system architecture and discuss an initial implementation using a portal as an example Web application.