2
edits
Changes
→Library News - A gathering place for library and tech news, and more
Find more information about Library News at the [http://news.librarycloud.org/faq Library News FAQ]
== Data-Mining Repository Contents to Auto-populate Scholarly Research Repository Submission Metadata ==
* Mark Diggory, Head of U.S. Operations
The existing body of Open Access scholarly research is a well classified and described dataset. However, in Institutional Repositories it can be the case that there are insufficient resources to invest for cataloging and maintaining rich metadata descriptions of contributed content. This is especially the case when collections are populated and maintained by non-librarians. A great deal of classifiable detail preexists within files that are submitted to scholarly repositories. Utilizing existing Open Source technologies capable of extracting this information, a process can be provided to submitters and repository maintainers to suggest appropriate subject classifications and types for descriptive metadata during submission and update of repository items. This talk will provide an overview of an approach for utilizing machine learning as a tool for the auto population of subject classifications and content types.
[[Category: Code4Lib2012]]