User:CMcWebb
From Open Annotation Collaboration
Contents |
Christine McWebb
Associate Professor
Associate Chair, Graduate Studies
Département d'études françaises
University of Waterloo
Ian Davis
Freelance Consultant & Lead Developer of MAT (MARGOT Annotation Tool)
University of Waterloo
Use Case: MARGOT Annotation Tool
Description
What
The MARGOT Annotation Tool (MAT) is a one of the sub-projects of MARGOT (Christine McWebb, PI), which was founded in 1993 and is housed and maintained by the University of Waterloo. As a testbed for the tool prototype development (January 2011-December 2011), we will initially use the small dataset of 100 medieval manuscript miniatures of the well-known French Roman de la rose jointly written by Guillaume de Lorris and Jean de Meun (1236 and 1269-78). These images have been made available in high resolution scans on MARGOT and are an integral part of McWebb's "Reading the Roman de la rose in Text and Image" project, complementing her critical anthology Debating the Roman de la rose (New York: Routledge, 2007). A great number of the over 300 extant manuscripts are richly illuminated and, consequently, provide a useful dataset for scalability. As a starting point, however, this small dataset lends itself well to the development of the functions that make up the tool we will create because the images are relatively homogeneous, as they all stem not only from the same literary work but also from a defined number of excerpts from Jean de Meun's continuation and from chronologically limited manuscripts spanning a period of about seventy years in order to match the chronology of the critical anthology. This step in the development will show the value of the tool for large data repositories and will turn our collection of Rose miniatures into an interoperable repository of images targeted for annotation.
Rather than using a large image repository for the initial development, a small dataset will be more beneficial at this stage in order to provide a set of well-defined visual characteristics, such as objects, colors, illuminations, shapes and so forth. Further, a small dataset will make backtracking easier when programming and coding errors need to be located, and it will facilitate the task of indexing the terms used in the annotations. However, our software will be designed to be scalable to allow for application with large datasets. In year 2 (2012), user testing will be carried out using the following existing archives: The Digital Library for Medieval Manuscripts at Johns Hopkins University, e-codices at the University of Fribourg, Switzerland, and Parker on the Web, Stanford University.
Scope
The screenshots below demonstrate a range of possible annotations that would be carried out by scholars, students, repository editors (see under "Who"):
Screenshot 1 shows examples of possible annotations that can either be made public on the repository site or kept hidden from public view.
In Screenshot 2, we demonstrate how the user will be able to retrieve for example all occurrences of the same allegory, in text and image, or even of a component (a piece of clothing, a color etc.) of that same allegory, in other words we show the aggregation capabilities of MAT.
Who
The different types of users we envisage for the tool are these: a project staff member who has the task of creating annotation content (for example, an art historian employed by ARTstor who needs to annotate a set of images for public viewing, as demonstrated in Screenshot #1); a member of the broader community who wishes to annotate a set of images for his/her own private usage; a member of the broader community who wishes to share his/her annotations on the site of the image repository with the caveat that quality standards are met via a workflow approval process through authentication/authorization.
Details
Relevance
MAT will be a web-centric application grounded in accepted design standards that are currently being developed by related projects led chiefly by the Stanford University Libraries and the Open Annotation Collaboration (OAC). This collaboration will ensure that MAT's technical design will allow it to be interoperable from the early development phase forward and that we build a user-friendly tool that produces first-rate data for consumption. We will be able to benefit from a community of developers working toward a common goal, which is to provide standardized web-services for scholars interested in annotating images in online repositories.
Challenges
The 2005 Summit on Digital Tools in the Humanities reported that "only about six percent of humanist scholars go beyond general purpose information technology and use digital resources and more complex digital tools in their scholarship."[1] One would assume that this percentage has increased in the last five years, as researchers, including humanists, increasingly turn to the web as a starting point for their research in order to access these materials. Yet, these findings have most recently been confirmed by a preliminary evaluation of a survey titled "Humanities and Social Sciences Scholars’ Use of Digital Technology for Teaching and Research" conducted by INKE (Implementing New Knowledge Environments, University of Victoria, Canada).[2] It is perhaps not surprising that the percentage of digital tool users has remained low because, thus far, most digital tool development adhered to project-driven design principles, where the goals of a project guided the software development. The reverse strategy, where software development becomes the focus and not a later addition to a scholarly digitization project, has been adopted by such projects as IMT (Image Markup Tool, Martin Holmes, University of Victoria), IBX (Kevin Kiernan, University of Kentucky, emeritus; Emil Iacob, Georgia Southern University, formerly EPPT: Edition Production and Presentation Technology), designed to integrate images and text through XML, to make them available for use by any image-based electronic project, and Pliny (John Bradley, King’s College), which is a prototype for free downloadable software that facilitates note-taking and annotation. With the MARGOT Annotation Tool, we propose to follow the lead of these projects, while at the same time addressing the overriding weakness that these tools share: limited interoperability.
References
- Summit on Digital Tools for the Humanities, 2006, http://www.iath.virginia.edu/dtsummit/SummitText.pdf.
- Stéfan Sinclair presented these findings at the Digital Humanities Summer Institute, University of Victoria, June 7-11, 2010.
