Towards a Norwegian general thesaurus? Unni Knutsen, Humanities and Social Sciences Library Mari Lundevall, Science Library.

Towards a Norwegian general thesaurus?

Unni Knutsen, Humanities and Social Sciences LibraryMari Lundevall, Science Library

Subject indexing and classification

Report from working group (2010):• Lack of coordination across (and within)

faculties: – Uncontrolled index terms, controlled index terms,

two thesauri: Humord (humanities and social sciences) and MESH

– 7 different classification schemes (including DDC and NLM Classification)

• Resource constraints

Recommendations

• Use established subject indexing systems• Extend Humord to include most of the

subject areas in the library– In addition: MESH

• Do away with in-house classification schemes, use DDC and NLM Classification

More about Humord

1. A thesaurus (humanities and social sciences)

– 26 000 concepts

2. In addition the name of a joint indexing activity within the framework of the shared catalogue – BIBSYS– Cooperation and reuse of indexing data– Consistent use of indexing terms based on

common indexing rules

National Library of Norway

• Report from working group:– Various in-house subject indexing systems– Lack of coordination – Lack of standardized subject indexing systems

• Tested data from various sources. Conclusion: Use of Humord gave the best result

So we joined forces…

• The University of Oslo Library received funds for 2014 to:1. Participate in a study with the National Library to

explore the feasibility of developing a more general, national thesaurus based on Humord and the controlled vocabulary for natural sciences and mathematics

2. Develop methods of mapping from Humord to WebDewey

Realfagstermer

• Controlled vocabulary for natural sciences and mathematics

• 14 000 concepts– Synonym control– Related terms– English (10%), some Latin– Subject strings

Mapping attempts

• Target: the on-going Norwegian DDC-translation

• 500-group, 600-640• Computer assisted direct matching: Mapping

suggestions based on string matching from our terms to DDC captions

• Co-occurrence mapping: Mapping suggestions based on co-existing DDC and subject terms in catalogue records

Term in source vocabulary

Term in target vocabulary

μmapper Example