3 Olga Murdoch Lorcan Coyle Simon Dobson 2008 Derek Bridge, Ken Brown, Barry O'Sullivan, and Humphrey Sorensen 19th Irish Conference on Artificial Intelligence and Cognitive Science Cork, Ireland 103-112 27/08/2008 ontology, recommender systems, query elaboration, image retrieval, image annotation, tagging Stock photo libraries are the most common means for publishers and advertisers to find images for their media. Searching for the perfect photo can be a time-consuming and frustrating task. This is because searching is often dependent on the descriptors or tags given to each photo by the editors and contributors to the library. The tagging process is subjective, further complicating the search process. We describe an algorithm that uses domain ontologies to improve the interactions with these libraries. Ontologies are used to expand query terms based on users' initial search queries. We present results that demonstrate that the use of ontologies greatly improves users ability to retrieve photos when undertaking a number of search tasks.