In lieu of an abstract, here is a brief excerpt of the content:

11 Cyberenvironments Ubiquitous Research and Learning james d. myers and robert e. mcgrath Over the past fifteen years, the World Wide Web has evolved from a tool created to support scientific research to a ubiquitous social infrastructure. The Web has had an enormous impact on society in terms of changes in practices and culture, in the emergence of new businesses and career paths, and in the extent to which our lifestyles have become dependent on its existence. Yet a mechanistic description of it—that it greatly simplifies sharing of text and multimedia information and enables links between documents—provides no more than a hint of its transformative power. Nonetheless, the Web has revolutionized our notion of how information is created and by whom. Wikis, blogs, and related tools enable personal publishing and the community development and evolution of resources (Wesch 2007), in stark contrast to traditional authoring and publishing. Who would have imagined when the first browser appeared that we would be able to access encyclopedias created by thousands of individuals (e.g., Wikipedia), and further, that we would be able to contribute our own knowledge back to them, from our home computers, in real time? In science and engineering, adoption of the Web has been just one part of the broader adoption of cyberinfrastructure—high performance computing, instruments, data, networks of sensors, and analysis and visualization services all available via the Internet (Atkins et al. 2003; National Science Foundation Cyberinfrastructure Council 2007). Riding on the exponential increases in computational , data storage, and networking technologies, the development of “grids”, “e-science”, “science gateways”, and “community databases,” has given researchers access to more, and more powerful, resources than ever before, with a direct effect on scientific productivity. 120 . myers & mcgrath However, we believe that cyberinfrastructure is also beginning to foster change in our conceptualization of research and learning processes in ways analogous to the Web’s impact on our notion of information authoring and publishing. The availability of directly accessible data, instruments, and computational resources supports and is helping to catalyze a shift in scientific research toward multidisciplinary, systems-oriented studies and to close coupling of computational modeling with experimental observation. Researchers are also increasingly publishing data and experimental procedures independently, in addition to writing papers to summarize their work, which is enabling their colleagues to quickly assemble the data, instrumentation, and computational resources needed to reproduce results quickly and to extend the work in new directions. Similar to the way the Web enabled linking across Web sites, scientists can now link instruments, models, and analyses in computational workflows. Instead of a creating a text-centric encyclopedia, researchers are coordinating at the community level to collaboratively create reference databases that can feed data directly into workflows. The National Center for Supercomputing Applications (NCSA) has coined the term “cyberenvironments” to describe systems designed to support both traditional research activities and the evolution of these types of new collaborative practices. Rather than focusing solely on access to advanced resources, cyberenvironments emphasize the continual creation of new resources, the dynamic integration of these shared resources into projects, and the direct publication of the new resources created in projects back into the community-level scientific context. Cyberenvironments are intended to support researchers in efficiently discovering, accessing, and integrating resources to explore new ideas and in disseminating their work, in a detailed and actionable form, to their colleagues. How does this impact learning? First, working across disciplines and making use of shared resources requires much more just-in-time, agile learning and emphasizes packaging of resources with all the information needed to understand them (using technologies from the Semantic Web, for example). Thus a model made available as a service will link to the paper in which it is described, to sample data and outputs, to analyses in which it has been used, to notes about its quality and applicability for certain problems, to alternate models, and so forth. Knowing who created, annotated, and used the model in turn provides connections to a network of experts who may be able to answer questions. As researchers couple their own work with models or data created by others, and the type of rich metadata described above becomes available, researchers will increasingly view learning as a continual process that occurs as one works rather than a prework exercise (or something that ends with formal schooling). Such a model becomes very scalable once seeded, as each researcher’s explorations add new metadata and new resources to the mix. [18.227...

Share