A thesaurus-based semantic library catalogue

As my master thesis I built a thesaurus for KABA controlled vocabulary used in Polish libraries. The work consisted of following steps:

analysis of KABA subject heading language,
detection of errors in KABA dictionary,
implementation of representation of KABA dictionary in Java, in order to make possible conducting computer analysis of subject headings stored in KABA dictionary,
transformation of KABA controlled vocabulary to a thesaurus. It was done by adding missing hypernymies. Fortunately, almost all of them follow from KABA subject heading language. This step required using simple natural language processing.
the design of following semantic applications of the thesaurus in library subject catalogues: thesaurus browser, semantic book search engine and statistical analysis of book subjects.

The goal of my work was to apply the idea of the Semantic Web to libraries. KABA is Polish counterpart of LCSH. It seems that KABA thesaurus can also be used to describe web pages' subject or even subject of individual paragraphs.

Fragment of the thesaurus containing vocabulary related to the police
KABA controlled vocabulary was translated to LCSH. Hypernymies are marked by lines. Thick ones mark hypernymies that are present in original controlled vocabulary. Thin labeled lines mark hypernymies added as a result of master thesis.

Semantic book search engine
Thanks to thesaurus it is possible to find all books about policemen i.e. about policemen, their training, ethics etc. All books that we are interested in can be returned in one common list. The most popular books will be returned at first place.

Documents:

Master thesis summary - presented at 3rd Language & Technology Conference, Poznań, 2007
Description of the idea at "Biblioteka 2.0" ("Library 2.0") forum - in Polish
Master thesis text - in Polish