Use Case Pode
Back to Use Cases & Case Studies page
Anne-Lena Westrum (project leader), email@example.com
Biblioteklaboratoriet (The library laboratory) and the Norwegian authority for archives, libraries and museums.
Background and Current Practice
The purpose of the Pode project has been to use technology and external data to enrich the data in library catalogue records, and thereby creating a platform for developing enduser services with information and functionality that is not available in the current library web search. The actual development of light applications to demonstrate the use of data has been included in the project.
The first phases of the project has concentrated on using library protocols to access bibliographic data and mashing these up with data available from non-library data sources through web service APIs. The current phase of the project concentrates on converting library data to RDF and linking data to individual instances in other LOD datasets.
This phase of the project consists of two tasks. Described separately:
A1. Presenting the productions of two authors grouped by FRBR entities Work and Expression, enriched with outside information about these authors and links to online fulltext versions of books.
A2: Converting FRBRized bibliographic data to RDF, and enriching these data with links to individual instances in DBpedia, VIAF and Project Gutenberg.
B1. Making an application that allows the enduser to navigate related Dewey categories by labels in several different languages, to browse the library's collection of multilingual non-fiction documents.
B2. Converting MARC records to RDF. Creating links to Dewey categories with multilingual language labels from http://dewey.info/.
The library and library science community. Developers that want to use library data in web services.
Use Case Scenario
The first task has been to export the catalogue records connected to two Norwegian authors, Knut Hamsun and Per Petterson. The catalog records have been run through an automated FRBRizing process, creating instances of expressions and works for the collection, and relating these to each other. Furthermore the FRBRized data have been converted to RDF turtle syntax, and we have enriched those data further by adding links to instances in the DBpedia, VIAF and Project Gutenberg datasets.
We are currently working on a web application that uses these data to let the enduser browse the authors' production grouped by the abstractions work and expressions, as well as adding extra relevant information from other datasources.
Also we have started working on a new task, in which we aim to convert the catalogue records for the whole collection of non-fiction literature at The Multilingual Library in Oslo (ca. 16 000 documents in more than 90 different languages). The idea is to let the user browse the collection by Dewey categories, by using the linked data representations of the top levels of DDC with multilingual labels from http://dewey.info/ .
Bibliographic data have been converted to RDF and enriched with links to external instances that describe persons and works relevant for the dataset. These links are used to get information that supplement the bibliographic data and that can be used to add extra functionality.
Related Vocabularies (optional)
- Core FRBR
- Dublin Core metadata terms
- Bibo (Bibliographic ontology)
Problems and Limitations (optional)
FRBRizing catalogue records demands a high degree of consistency in the cataloguing practice, to get a good result. This is usually not the case; cataloguing practice tends to differ between different departments, cataloguers and times. In order to get a result as good as possible, we had to correct several records. An attemt at FRBRizing the whole library collection would not give an output with the same quality.
The linked data version of Dewey decimal classification is so far limited to the top three levels of the classification system. This does not provide the necessary granularity for some classes.