With generous support from the Old York Foundation and The Durst Organization, Avery Library engaged in a three-year project to inventory, catalog, preserve and digitize the Old York Library collection and make as much material as possible available online. The project honors Mr. Durst’s important tradition of collection-building and facilitates active use of the collection in curricular and research activities, at the same time it provides online access for public audiences interested in New York’s history and real estate development.
The project team engaged in three main areas of activity: data development, digitization, and website development.
The multi-format nature of the materials required a multi-faceted approach to achieving intellectual control over the collection objects, and to develop rich metadata that would support the online discovery environment we envisioned. We developed a custom metadata schema that takes advantage of several standards including: MARC for bibliographic items, MODS-derived custom metadata schema for non-bibliographic items, EAD for archival materials, and VRA Core extended for visual materials.
We utilized a combination of enterprise platforms and custom-built project database platforms to inventory and catalog the collection, to preserve the data and digital assets, and to make them accessible to online discovery. Our database environments include: DIMS (locally-developed custom inventory management system), Voyager (network cataloging platform), Hyacinth (metadata editor), and FEDORA (our institutional repository).
A significant goal of the project was to not only to inventory and create accessible online records for all items in the Old York Library collection, but also to present as many digitized copies of those items as possible, online. Our decision to do so prompted several efforts: one to establish a policy for interpreting copyrights and fair use guidelines as they applied to digitization of materials for the project website; another to identify which items had been digitized by others and were available in high resolution on other open access sites; and thirdly to perform physical examination of all the items in the collection to determine if they could safely be digitized without harming the original object. Through this process, we established four different digitization workflows, two of which employed contracts with external digitization services, and two of which utilized internal digitization labs at Columbia. As a result of our analysis processes, more than 20,000 digitized items can be discovered and viewed via the collection website.
All digitized items are discoverable through the Seymour B. Durst Old York Library website. Digitized items hosted on other sites are linked to from the data records, these include links to items archived in Hathi Trust and the Internet Archive. Non-bibliographic items are also viewable in the Columbia University Libraries Digital Library Collections website. In addition, items from the collection are queued to be shared with Digital Public Library of America (DPLA) and are optimized for discovery by standard search engines such as Google.
A chief goal for the website project was to enable access and use of collection items in an openly accessible online environment. The site offers multiple discovery pathways for exploring the collection: search by keyword or mapping interface, browse all items or browse for pictorial materials (images) only. All search and browse results can be further narrowed using our faceting feature, i.e. users can narrow their search/browse result sets by location (city, borough, neighborhood) or by item format (books, pamphlets, maps, postcards, etc.). In addition, search/browse results can be sorted by title or relevance, displayed in variable item-lists per page, or in a light-table grid.
Other features of the site were developed to make collection items most usable. Visual materials can be examined online using our richly featured image viewer or downloaded in multi-resolution JPG format. Item records can be exported in XML format, and citations per item can be auto-generated in APA, Chicago or MLA citation format. The site also enables sharing links to records via a wide range of social media tools (such as Twitter, Facebook, Pinterest, and LinkedIn, among others).