The Ohio Library and Information Network (OhioLINK) is a consortium of college, university, and public libraries in Ohio. Its mission is to provide Ohio students, faculty, and researchers with the information needed to advance teaching, learning, research, and the growth of Ohio’s knowledge-based economy. Each week more than 50,000 newly published articles are loaded into its Electronic Journal Center (EJC). OhioLINK users download 10 million articles a year.
The legacy system hosting the Electronic Journal Center was showing its age. As data grew, performance decreased.
User experience was also in need of an upgrade. By the time this project began, the EJC had contained approximately 22 million flat files that did not support full-text search. Search results were limited to the name and issue of the journal, along with a link to the PDF. Users were required to download the PDFs to find what they needed.
The system also had administrative limitations: Statistics needed for curation and purchasing decisions were derived primarily from server-side logs as there was no real way to track the materials accessed.
Three main applications were created with slightly different scopes. The first two were customer-facing, and the third, administrative.
1. The search interface to find the journals more easily and view them without necessarily downloading them for greater ease of use.
2. The “MyEJC” space for login and search saving enables the user to put journals on a “digital bookshelf.”
3. A member portal which is an internal back-office for the OhioLINK business user. It enables them to manage some of the metadata and other data going into the EJC.
The consortium teamed up with Insum, which, although initially unfamiliar with library systems, completely redesigned the user interface and back-end applications to specification. Insum also added powerful search functionalities compliant with international library standards (COUNTER). This last element was critical to the consortium’s funding model.
To save time, it was decided that the new platform would be run on an Oracle database as OhioLINK already had an Oracle installation base and well-qualified Oracle developers. The organization had also already developed several Oracle APEX applications and so APEX was retained as the development platform.
Insum developed new APEX applications and also chose Oracle Text to build out the back end of the search tool.
Oracle Text is a feature in the Oracle database that allows typical “Google-type” searching such as full-text search. It a perfectly appropriate tool for such large amounts of data. It can full-text search PDFs, OCR content and it can also index word document supported formats.
Oracle Text is advanced. It has natural language processing capability: It can search for “sound like” items, deal with small nuances of language. It also supports “wildcard” searches, where putting an asterisk before or after a search word also pulls up its derivatives.
To enable users to deepen their search, a key Oracle Text feature used in the EJC project is « faceted browsing » which displays the categories the search word belongs to. As EJC users view their search results, they can check off categories and thus filter down their search to get a manageable set of results.
Another important feature OhioLINK required was advanced search. This allows EJC users can remove the fields they don’t want. They can also choose Boolean type searches and choose to exclude terms. When they are ready, they can perform a more restricted search at the touch of a button.
OhioLINK now has a stable platform with significantly enhanced performance requiring less operational support. It continues to host its vast resources through greatly improved search and reporting capabilities while provide unrivalled research capabilities to its customer base.
-Jeff Smith, Director of Shared Infrastructure at OH-TECH
-Judy Cobb, Manager Digital Platforms, Ohio Link