Annotation Supporting Collaborative Development of Scholarly Editions

Partner Insitution: The University of Queensland, Australia

This collaboration with AustLit evaluates and demonstrates the applicability of Open Annotation in the context of annotating variations between versions of a literary work. This experiment explores annotation of annotations (including nested annotations); annotations involving multiple targets and/or ORE aggregations as targets; annotations involving structured text as annotation body (content); distinctions between reference and targeting; and issues arising with tools that enable annotations spanning multiple manifestations of FRBR expressions and works. It also demonstrates the potential for robust retrieval by target of annotations described using the OAC data model.

Queensland Annotation Experiment Final Report

Annotation of Digitized Medieval Manuscripts

Partner Institution: Stanford University

This collaboration with the Stanford-led effort to define a modular and interoperating environment for digitized medieval manuscript collections allows Open Annotation to explore annotation of both structured text and still-image media, implying also a broad range of use cases and classes of annotation, e.g., from granular (evidence on a page image of a scribe's hand) to broadly scoped (on the cultural significance of a particular graphic construct used). This work will inform Open Annotation regarding annotation of annotations (e.g., annotation of transcribed text annotations); multiple approaches to and tools for segment description (see also below regarding TILE and T-PEN initiatives), including trade-offs between minting new URIs and providing segment descriptions; annotation bodies that are structured (e.g., TEI); annotations of aggregate targets and/or multiple targets, some of which may be discontinuous and/or distributed across repository boundaries (e.g., Stanford's "return-and-reuse" scenario, which can lead, for example, to a page image in one repository linked to a structured TEI transcript in another location); annotations that may change over time (e.g., need to update transcription annotations), providing an opportunity to examine both degree of annotation mutability supported by OAC data model and the potential importance in this domain of being able to exploit the Memento Framework to preserve the meaning of annotations that target mutable annotations.

Stanford Annotation Experiment Final Report

Annotation of Subscription Streaming Video Content

Partner Institutions: MITH & Alexander Street Press

Through this collaboration with the Alexander Street Press, Open Annotation is exploring sophisticated annotation of streaming video, focusing on issues to do with annotation target segment description for streaming media; the persistence vs. variability of such segment descriptions; and mixed media multiple target annotations (e.g., as when comparing and/or contrasting a segment of a performance with the corresponding segment of digitized text play script or music score). In many cases, scholarly content on the web not be freely available at open URIs, but scholars may still wish to annotate such content and link it to other content that is freely available. The Open Annotation data model must be supple enough to allow this. This collaboration allows us to examine issues to do with annotations having multiple targets spanning both subscription and freely accessible content. Finally, the Press already provides a tool to annotate clips of videos in their collection, but it does not currently provide a means to further annotate segments of the frame. Further, the annotation format used is specific to the interface, and thus cannot accept annotations from other sources or export its own annotations to external tools. The existence of an independent, domain and collection-specific annotation tool will suggest functional requirements and more importantly allow us to examine ways to import and export OAC-described annotations. The code developed by MITH for this project can be found at: https://github.com/umd-mith/OACVideoAnnotator. For more information about this project contact MITH project team members: Neil Fraistat, Jennifer Guiliano, Jim Smith, and Grant Dickie.

MITH Annotation Experiment Final Report

Annotation of Digital Emblematica

Partner Insitutions: University of Illinois at Urbana-Champaign & Herzog August Bibliothek Wolfenbüttel

Through this collaboration with the Emblematica Online project, itself a collaboration between the University of Illinois and the Digital Library of the Herzog August Bibliothek (HAB) in Wolfenbüttel, Germany and funded jointly by the NEH in the U.S. and the DFG in Germany, Open Annotation is exploring a broad range of use cases and classes of annotation involving both structured text and still image annotation targets. This work is informing the Open Annotation regarding annotation of annotations (e.g., annotation of transcribed and normalized mottos as annotations); annotations involving targets at multiple levels of granularity (e.g., motto, pictura, emblem, and book all potentially considered at item, manifestation, expression, and work levels); annotations having dissimilar targets and/or bodies (e.g., a Web accessible picture of an architectural emblem feature annotated by the motto of the print emblem instance from which the feature derived); annotations having annotation bodies drawn from a hierarchical controlled vocabulary with persistent identifiers (i.e., the multi-lingual Iconclass Thesaurus), thereby exploiting the separation of Annotation from Annotation Body inherent in the OAC data model. Emblematica Online has implemented a prototype emblem registry shared by UIUC and HAB and plans to expand this registry to include emblem books. CNRI-style handles are utilized. This collaboration provides an opportunity to test the utility of these registry and identifier services to support interoperable annotations.

Illinois Annotation Experiment Final Report

Maphub Phase II

Partner Institution: Cornell University

Cornell Information Science has released a first demonstrator of Maphub, which is a Web portal for georeferencing and annotating digitized, high-resolution historic maps. All annotations are represented in the W3C Open Annotation model and exposed as Web resources that can be accessed by various clients simply by dereferencing HTTP URIs. The portal has been bootstrapped with approximately 6000 public domain maps taken from the Library of Congress Historic Map division.

Here is a screencast introducing the central Maphub features. The Maphub Open Annotation API document explains and illustrates how Maphub implements the Open Annotation model. All system components are Open Source and available at http://github.com/maphub.

.

Cornell Annotation Experiment Final Report and Final Report Addendum

An OAC Compliant Annotation Framework in Fedora

Partner Institution: Brown University Library, Center for Digital Scholarship

The Brown University Library is developing an annotation service for Fedora using the Open Annotation data model. This project provides users of Fedora Commons digital repository software with add-on tools for creating, querying, and retrieving annotations based in Fedora objects and collections. While this software is intended to be usable with any Fedora instance, the investigators at Brown hope to interlink annotations built with their annotation experiment directly to the TEI-encoded texts contained in their Fedora repository in order to explore how annotations can be targeted to words, structural elements, and semantic data encoded within textual documents. This work will feature items from the Virtual Humanities Lab and Brown University Women’s Writers Project, as well as other collections of digitized texts used by the students and faculty at Brown. The code and documentation for this service are available at http://brown-university-library.github.com/oac_web_service/.

Brown Annotation Experiment Final Report

CODA - CATCHPlus Open Document Annotation

Partner Institution: The Meertens Institute

The Netherlands-based CATCH and CATCHPlus programs are building an OAC compliant annotation repository, document annotation tool, workspace services, and associated softeware tools using two bodies of existing annotations that have been scanned. Using tools developed during the course of the CODA project, the investigators at Meertens will produce OAC compliant annotations based on annotations of materials in the Queen's Cabinet collection at the Dutch National Archives and the Sailing Letters collection at the National Library of the Netherlands.

Meertens Institute Annotation Experiment Final Report

Automated Annotation of Biomedical Text

Partner Institution: University of Colorado Denver

Investigators at the University of Colorado Denver are adapting the Open Annotation data model to the context of automated annotation of biomedical literature, with the end-user focus of knowledge-based data analysis. To achieve the adoption of the OA model within their existing biomedical natural language processing and biological data analysis frameworks, the investigators at Colorado are extending the OA data model to accommodate several new requirements. Specifically, they are formalizing and testing new proposals for capturing biological relations and events as complex structured annotations (structured bodies), and modeling of provenance relations between annotations. The investigators at Colorado hope to demonstrate the feasibility of mapping from their automated natural language processing pipeline, built in the Unstructured Information Management Architecture (UIMA), to the extended OA-based data model. They hope to integrate the annotations with the Hanalyzer biological data analysis system under development at the University of Colorado School of Medicine, in order to support knowledge-based exploration of biological data by their primary target user group: biomedical scientists. In addition to biomedical annotation work, the investigators at Colorado are also trying to relate the OA data model to the Linguistic Annotatoin Framework. A recent paper on this work is available at: http://faculty.washington.edu/fxia/LAWVI/proceedings/cdrom/pdf/LAWVI10.pdf.

Colorado Annotation Experiment Final Report

Annotation Middleware for Scholarly Publications and Resources

Partner Institution: New York University Libraries

Investigators at NYU are building a working prototype of a stand-alone annotation web service that will work with content management systems, such as Drupal, that have robust commenting systems already built-in. Investigators hope to produce a prototype annotation system that can be used as a plug-in and provide more robust and portable annotations than existing plugins such as Commentpress, an existing annotation plugin for Wordpress.

New York University Annotation Experiment Final Report