ResearchSpace is made up of three fundamental and mutually supporting elements:
Collaboration – An environment which allows two or more people or organisations to work together to achieve common goals and objectives. This environment will employ the use of existing collaboration tools such as discussion forums, wikis, blogs and workflow, and be enhanced by the use of research tools (below) and other general productivity applications.
Harmonised data – Collection and conservation data held in a semantic form so that it can be shared and reused between projects without expensive integration software. This is described in the, ‘ResearchSpace Conceptual Framework’
document1. The British Museum is currently running a pilot of the technology to enhance their current Collection Online system with conservation and scientific data, and the National Gallery’s Raphael Research Resource also makes use of
semantic data.
Research tools – The provision of components that can be used independantly
or with other components for data analysis and manipulation to support a
particular collaborative workflow. For example, where a workflow requires the
comparison of two images of the same object processed using two different
techniques (perhaps an x-ray and a normal digital image of the same object), a
component could be available for overlaying the images, highlighting areas of
interest and allowing the researcher to record annotations.
This document describes the third element in more detail, including the environment
within which the tools would exist, their utility, and the relationship with the other two
ResearchSpace elements. The information is based on a ResearchSpace design meeting held in
December 2009, attended by representatives of all the Mellon prototype projects, and
subsequent preliminary analysis of the resulting requirements. These requirements will
ultimately provide a limited, definable scope for a ResearchSpace development proposal. It is
expected that requirements will continue to develop over time, and as such, the tools for
ResearchSpace would continue to grow and evolve to meet – rather than anticipate – the needs
of the research community.
Andrew W. Mellon Prototypes Two different types of prototype exist within the current portfolio of the Mellon
Museum’s program projects. The first type (e.g. The Raphael Research Resource at the National
Gallery, London and The Rembrandt Database, RKD) are more ‘results’ orientated and
concentrate on publishing existing data and/or data that is generated through research and
collaboration in venues unrelated to the prototype’s web presence. The results are presented
in a carefully designed web site, with the end goal of exposing information as reference
material for use by other researchers. Much of the work to bring the information together has
taken place away from the web site itself, and project personnel have utilised more traditional
methods of collaboration and communication. As such, this type of project faces some of the
barriers explained in the ResearchSpace ‘Conceptual Framework’ document.
The second type of prototype is represented by the project managed by the Courtauld
Institute of Art. Their objective in seeking Mellon funding was to test a new form of
collaborative, online research methodology for art historians, conservators, and scientists by
providing an environment for groups of researchers to generate ideas, debate, and discussion
around a defined set of questions. Rather than just present information generated elsewhere
as reference material, this prototype facilitates an ongoing, interactive, and iterative research
process that allows groups of scholars to interact and respond to the image and text data that
has been uploaded into the system, primarily through the use of a discussion forum. Such a
research process could result in users commissioning further investigation and analysis, or
possibly in publications. The Courtauld’s prototype has stimulated the thinking for many of the
proposed research tools required for the ResearchSpace environment.
The broad steps in the current Courtauld workflow are these:
- Identify the area of research. For example, a particular painting of interest by the artist, Maestro di Figline).
- Upload materials that support the research. This could be an image, a document or a video.
- Provide a forum to discuss these materials and focus the objectives of the research.
- Review, and if necessary, commission further work to go through the same process of evaluation by the collaborating experts.
- When ready and appropriate, publish the results.
This model is now recognised by others as a process that has universal application in
supporting research activity across a broad range of different subject areas. Most recently, the
Rembrandt Database team has identified that their immediate priority is to implement a
workflow similar to the Courtauld’s system in order to provide scholars with a more interactive
and dynamic space. It is their view that such functionality could assist with engaging other
holders of Rembrandt material and would ideally generate new information and
documentation to enrich their primarily presentational system.
The Missing Elements Although the workflow represented by the Courtauld site is generally universal and
provides a good foundation for collaborative research, it has two areas of weakness that
ResearchSpace could potentially resolve. The first is the lack of online tools available for
researchers engaging with the site. Currently, many processes and techniques still need to be
conducted away from the virtual environment and be fed back at a later stage (e.g. image
manipulation, overlay, enhancement and annotation, etc.). Such points of disconnection raise
the risk for a lack of consistency and continuity, both in terms of collaboration and data. Tools that facilitate online, real-time data interaction and manipulation constitute the majority of the
functional requirements identified by the ResearchSpace stakeholders at the December 9-10,
2009 meeting. Some of these are explored in greater depth in section 4.3 below.
Second, there is no consistent and managed data store for the information generated
on the site. The Courtauld’s system is essentially a content management system, a system
designed to simplify the publication of web content to a web site. As such it is supported by a
database designed to support this particular purpose and does not provide a system for storing
collection and scientific data in such a way that it can be managed for other scholarly purposes.
Even if the content management system was additionally supported by a collection or
conservation database system, as already discussed that system would not be compatible with
other systems used by other projects.
The red text in the representation of the Courtauld’s system below highlights the
inclusion of a small sample of tools that would address the identified desirability to keep the
activity and discussion within the virtual space. Such functions would facilitate more productive
collaboration and would enable data to be stored directly in the environment for others to use,
whether within the immediate project or elsewhere on ResearchSpace.
Look and Feel In trying to move away from the more traditional presentational model, it is important
that ResearchSpace not be conceived of as simply another form of content management
system concerned with web site production2. The toolset in ResearchSpace would include a set
of components or plug-ins that can be, ‘mixed and matched’ to support the research activities
and analysis required for a particular set of objectives. This model is not new: group
collaboration applications such as Microsoft SharePoint (a wide ranging collaboration
environment, similiar to Sakai) allows communities of developers to provide individual
standalone components that can be loaded into any of the collaborative spaces to enhance the
basic environment.
ResearchSpace would take the same approach but would address, in priority order, the
components and workflow that its stakeholders have identified as important for their work. It
would also have access to the underlying, ‘harmonised data’ environment mentioned above. In
this way researchers could configure for themselves the specific tools relevant to their
particular research agenda. Many of these components could also be made available to the
users of a published site so that visitors could perform their own analysis, thus providing a
wider, interactive engagement with the project’s research materials.
The Current Toolkit Requirement The toolkit, as currently defined, divides broadly into the following groups;
Data and image management – These are components that would allow a
project to organise and store data with the appropriate level of security.
Data and Image manipulation – These tools would allow transformation,
transcoding (the ability to convert between different file formats) and
generation of data.
Data and image analysis – These tools would allow searching, correlation,
comparison, annotation, definition of relationships, etc.
Publication and presentation – These tools would allow data to be published in
a web format.
Tools of this kind, when added to the collaborative workflow model established through the
Courtauld prototype, transforms the environment into a far more robust, interactive system
which can create, facilitate, store and publish research information and data more quickly.
The full list of the individual tools, logically grouped and prioritized for phased development,
would be part of the pilot phase of ResearchSpace and addressed in full in a proposal for such a
pilot.
Next Steps The requirements collected in the December 2009 meeting require sorting into logical
groups and placed in priority order.This undertaking should include the requirements gathered
in the ConservationSpace project design phase so that overlaps in web functionality can be
rationalised. This list of requirements should be compared with other tools already developed
by the open source community (particularly those developed, or currently being developed, in
other Mellon initiatives) that could be reused in the proposed ResearchSpace environment.
Additional workflows to the one represented by the Courtauld prototype also need
identification, consultation with stakeholders, and must be documented as part of the
ResearchSpace specification.
It would also be possible to conduct a pilot that would demonstrate the Courtauld’s
workflow enhanced by the development of some of the proposed research tools. This pilot
could also take the work of the National Gallery and British Museum a step further by creating a
harmonised semantic data store, but with a much larger sample of collection and conservation
data, demonstrating the scalability of the technology.
Dominic Oldman, Feb 2010
|
