|
 back
|
|
Building a Digital Library for Public Health Professionals
Elizabeth D. Liddy, Center for Natural Language Processing, School of Information Studies, Syracuse University, Syracuse, NY, USA
Abstract:
Much of the useful information in public health (PH) is considered grey literature, i.e. literature that is not available through traditional, commercial pathways. The diversity and non-traditional format of this information makes it difficult to locate and collect, and also a rich domain with high potential payoff for the development of a Digital Library. Under funding from the Robert Wood Johnson Foundation, we set forth to improve access to PH grey literature reports through established natural language processing (NLP) techniques. First, we needed to establish a model-based metadata schema for automatically analyzing and representing the PH grey literature. To accomplish this, we analyzed a corpus of grey literature from PH websites to determine a candidate set of metadata elements, followed by review by PH professionals who prioritized the elements for NLP extraction into a metadata record. The automatic NLP extraction technology was used to populate this schema for a collection of PH reports and evaluated by experts. Current research, funded by NLM, is building on these early successful results for the construction of a fuller digital library and an evaluation of its contribution to the work practices of PH professionals in county health departments in New York and Seattle.
|
|
|
