Share on facebook
Save on delicious
RSS Feed
A case study about how e-Infrastructure is used within the Social Sciences
A case study about the use of e-Infrastructure elements used within the Social Sciences
Mercedes Argüello Casteleiro1, Pascal Ekin1, Simon Peters 1, Michael Fraser2, Peter Halfpenny 1, Matthew Mascord 2
1 University of Manchester (UK)
2 University of Oxford (UK)
Mercedes.ArguelloCasteleiro@manchester.ac.uk
Introduction
- -----------------
Nowadays, the Internet lies at the core of an advanced scholarly information infrastructure, facilitating access to distributed data and enabling information intensive collaborative research. The deployment of e-infrastructure, whether within institutions, nationally or internationally, has the potential to increase the pace, impact, and efficiency of research both within and across disciplines. In the future, Internet virtual documents will be enhanced with virtual computing, based on existing and emerging e-infrastructure. This will shift research to a higher level at which information tools are encapsulated as services accessible on-line. Thus tools formerly available only to specialists will become accessible to all, with the consequence that current intensive manual or semi-automated data-processing and data-analysis will be automated through access to the appropriate services.
If academic research in the UK is to build on the foundations laid by UK e-Science then it is essential to understand the process by which different research communities adopt e-infrastructure, and use this understanding to ensure that the required interfaces, support and training are put in place to enable progressive adoption across all research communities. The eIUS project is a JISC funded project that aims to study current and planned usage of e-infrastructure across research communities. One objective of eIUS is to help to make the available and emerging e-Infrastructure more visible to the academic community and foster its use. A second objective is to help e-Infrastructure service providers to better understand how users engage with their services.
The eIUS project intends to accomplish its objectives through an iterative process of experience capture and analysis carried out in conjunction with the UK research community. A scoping e-Infrastructure usage report with the outcomes of the piloting work done for eIUS has been recently released. This paper focuses mainly on an output that the eIUS project will be producing: Service Usage Models (also known as SUMs) that describe patterns or combinations of e-Infrastructure services.
e-Infrastructure and service-oriented approaches
- --------------------------------------------------------------------
Although service-oriented approaches are already being applied successfully, in some cases at substantial scales, much more effort is required before these approaches are applied routinely across all the natural sciences, social sciences, and humanities. Grid technologies can accelerate the development and adoption of service-oriented approaches by enabling discipline-specific content to be separated from domain-independent software and hardware infrastructure. In order to capitalize on these opportunities, pathways for the diffusion and adoption of e-infrastructural innovations must be first understood and then made as smooth and well supported as possible.
The e-Framework is an initiative that has four international partners, including JISC. The primary goal of the initiative is to facilitate technical interoperability within and across education and research through improved strategic planning and implementation processes. In pursuing this goal the e-Framework intends to support, among others, a service-oriented approach to system and process integration and community involvement in development of the e-Framework. In supporting a service-oriented approach it has the potential to maximize the flexibility and cost-effectiveness with which IT systems can be deployed, in institutional contexts, nationally and internationally.
SUMs and CORE SUMs of the e-Framework
- ---------------------------------------------------------------
Service Usage Models (SUMs) are a core component of the e-Framework. SUMs provide a description of the needs, requirements, workflows, management policies and processes within a domain and the mapping of these to a design of a structured collection of Service Genres and Service Expressions, resources, associated standards, specifications, data formats, protocols, bindings, etc., that can be used to implement software applications within the domain. In other words, SUMs model how services meet business needs.
A Commonly Recurring SUM (CORE SUM) is a core e-Framework term referring to a description of the key aspects of a common design for an operation, behavior or communication used in creating a reusable service-oriented design. CORE SUMs are identified from examples, not created.
According to the e-Framework:
. A Service Genre is a collection of related behaviors describing an abstract capability that supports a business process. Service Genres are by definition generic, so they can easily be part of more than one Service Usage Model or CORE SUM.
. Service Expressions represent specific cases of the related but abstract and generalized Service Genre.
The eIUS project contribution to the e-Framework SUMs
---------------------------------------------------------------------------------
The e-Framework allows the various stakeholder communities to document their requirements and processes in a coherent way, and to use these to derive a set of interoperable network services that conform to appropriate open standards. By documenting requirements, processes, services, protocol bindings and standards in the form of Service Usage Models (SUMs), service providers are better able to collaborate on the development of service components that meet the needs of the research.
One way of contributing to the success of the e-Framework initiative will be to have a large number of Service Usage Models. This paper presents a SUM and two CORE SUMS that follow the e-Framework description and template for Service Usage Models (SUMs). The SUM is based on a Grid-based application for social sciences: GEMEDA; while the two CORE SUMs were obtained by identifying reusable e-infrastructure elements and patterns of use. The two CORE SUMs seem to be generic enough as to be valid not only within social sciences but also across disciplines; so far they have been successfully validated by means of another Grid-based application for social sciences: ConvertGrid.
GEMEDA (Grid Enabled Micro-econometric Data Analysis) is a Grid-based application for social sciences that integrates the data, computation and presentation elements of an empirical economic modeling process. Although the use of Grid technologies is currently not common within social science research, many social scientists today want to investigate complex research questions that mean combining datasets from a variety of sources. However, there are difficulties in converting and combining the various data. GEMEDA has provided invaluable insights into many of the technical and methodological issues that need to be addressed when Grid-enabling social science datasets and developing Grid-based services. ConvertGrid has a wider scope than GEMEDA and it is intended to contribute both to capacity building within social sciences, and development of an emerging e-Social Science Data Grid.
Before a technical component can be submitted to the e-Framework, it must be described and documented in a prescribed way. The Service Usage Model Description is the documentation used to provide detailed information about the set of requirements and processes that the Service Usage Model represents. The template of the Service Usage Model Description has been divided into sections where each section indicates briefly the information that should be provided; some of sections are optional.
On the one hand the paper uses the GEMEDA and ConvertGrid examples to show how reusable e-infrastructure elements and patterns of use are being identified within the eIUS project, and on the other hand the paper illustrates how the most relevant sections of the Service Usage Model Description can be easily populated. To illustrate the latter, the description section of the Service Usage Model Description is an informal, non-technical narrative describing what the SUM does and how it does it; based on the study of documents provided by the researchers and on their experience gathered through informal interviews that take them through their research cycle, the description section GEMEDA SUM includes:
An empirical economic modeling using secondary information is articulated by means of bringing together the following services
. Data handling service - provides the functionality to extract the required data
. FTP service - provides the functionality to transfer data between services (e.g. Data handling service and HPC service)
. HPC service - provides the functionality to perform the econometric computations
. GEMEDA service (correlation service) - provides the functionality to act as a gateway between the user requests and the rest of the services
. Authentication and Authorization service - provides the functionality to facilitate the user a single sign-on
Conclusions
- ------------------
A service-oriented approach has the potential to maximize the flexibility and cost-effectiveness with which IT systems can be deployed, in institutional contexts, nationally and internationally. Service Usage Models (SUMs) are a core component of the e-Framework that model how services meet business needs.
The eIUS project is designed to help e-Infrastructure service providers to better understand how users engage with their services by developing SUMs that describe patterns or combinations of e-Infrastructure services. However, despite the e-Framework efforts to develop templates, instructions and illustrative examples, as well as formal definitions, developing SUMs and CORE SUMs remains a difficult and time-consuming task. This paper uses the experience of the eIUS project to examine some of the difficulties and reflects upon methodologies that would facilitate the development of useful SUMs in a way that minimises the effort required of developers and end-users when gathering the relevant information from them and also motivate them to contribute actively to grounding SUMs in their experiences.
References
- ----------------
C. Borgman (2007): Scholarship in the Digital Age. Information, Infrastructure, And The Internet. Mit Press Ltd (United States)
G. Schreiber, H. Akkermans, A. Anjewierden, R. de Hoog, N. Shadbolt, W. Vand de Velde, and B. Wielinga (2000): Knowledge Engineering and Management: The CommonKADS Methodology, The MIT Press.
I. Foster (2005): Service-Oriented Science, Science, 308, pp 814-17.
ConvertGrid, http://pascal.mvc.mcc.ac.uk:9080/convert
e-Framework, http://www.e-framework.org/
eIUS, http://www.eius.ac.uk/
eIUS Scoping e-Infrastructure Usage, Interim Report, http://www.eius.ac.uk/scoping/
GEMEDA, http://www.ncess.ac.uk/research/pilot_projects/gemeda/
JISC, http://www.jisc.ac.uk/
