Mark Patton

Using the Gamera framework for the recognition of cultural heritage materials

This paper presents a new toolkit for the creation of customized structured document recognition applications by domain experts. This open-source system, called Gamera, allows a user, with particular knowledge of the documents to be recognized, to combine image processing and recognition tools in an easy-to-use, interactive, graphical scripting environment. Gamera is one of the key technology components in a proposed international project for the digitization of diverse types of humanities documents.

The Archive Ingest and Handling Test: The Johns Hopkins University Report

From very early in its existence, the Digital Knowledge Center (DKC) in the Sheridan Libraries at Johns Hopkins University (JHU) has focused on using automated and semi-automated processes to create workflows for the creation and ingestion of digital objects. What was missing was a place to put these objects, a standard way to put them there, and a way to preserve them. This has begun to change over the past two years.

Services for a customizable authority linking environment

The SCALE - services for a customizable authority linking environment - project is developing tools to help integrate collections within the National Science Foundations National Science Digital Library (NSDL) through terminological linking. These tools transparently provide reading support to NSDL collections by automatically linking phrases in the content of one collection to relevant material in another collection. This enhance the user experience in the NSDL and encourage greater integration of collections by increasing the value of each collection to other collections.

Enhancing access to the levy sheet music collection: reconstructing full-text lyrics from syllables

The goal of the Lester S. Levy Sheet Music Collection, Phase Two project is to develop tools, processes, and systems that facilitate collection ingestion through automated processes that reduce, but not necessarily eliminate human intervention[1]. One of the major components of this project is an optical music recognition (OMR) system[2] that extracts musical information and lyric text from the page images that comprise each piece in a collection. It is often the case, as it is with the Levy Collection, that lyrics embedded in music notation are written in a syllabicated form so that each syllable lines up with the note or notes to which it corresponds. Searching the syllabicated form of words, however, would be counterintuitive and cumbersome for end-users. This paper describes the evolution of a tool that, using a simple algorithm, rebuilds complete words from lyric syllables and, in ambiguous cases, provides feedback to the collection builder. This system will be integrated into the workflow of the Levy Sheet Music Collection, but has broad applicability for any project ingesting musical scores with lyrics.

Capturing and curating published data

Verifiability and reproducibility are core tenets of the scholarly communication process. For many scientific publications, however, it is often the case that supporting datasets are not preserved, even when the article text is. And when they are, it is usually as a collection of files without relationships amongst one another or to the articles with which they are associated. There are some existing approaches that attempt to link datasets with articles after the fact (e.g.,NED ), but they are relatively few and involve substantial human intervention. The Digital Research and Curation Center in the Johns Hopkins University Sheridan Libraries, in conjunction with its partners has developed a proof-of-concept system that demonstrates an approach to capturing datasets during the process of submitting the associated article. As part of this process, linkages are established between the datasets and the article.