As library collections increase rapidly, personalized recommender systems have become a very important service for library patrons. The purpose of this paper is to design and implement a personal ontology recommender (PORE) system by building personal ontologies based on patrons' borrowing records.
In the PORE system, the traditional cataloging scheme, classification for Chinese libraries, is used as the reference ontology. This reference ontology is transformed to a unique personal ontology for each user based on the mining results from library borrowing records of that user.
A personal ontology represents a unique user interest on specific subjects. The personal ontology can be used to filter out unsuitable recommendations based only on a keyword matching method. Besides, the recommended books can be organized into the personal ontology, and provide the patron with a user‐friendly interface to access library collections.
The PORE system is currently implemented for Chinese collections. From this paper builds a new version to support English collections by adopting the Library of Congress Classification as the reference ontology.
This paper represents a practical method of building a user's personal ontology and explains the functional use of ontology knowledge.
