1 Abstract��� Personalized content retrieval aims at improving the retrieval process by taking into account the particular inter- ests of individual users. However, not all user preferences are relevant in all situations. It is well known that human pref- erences are complex, multiple, heterogeneous, changing, even contradictory, and should be understood in context with the user goals and tasks at hand. In this paper we propose a method to build a dynamic representation of the semantic context of ongo- ing retrieval tasks, which is used to activate different subsets of user interests at runtime, in a way that out���of-context prefer- ences are discarded. Our approach is based on an ontology- driven representation of the domain of discourse, providing en- riched descriptions of the semantics involved in retrieval actions and preferences, and enabling the definition of effective means to relate preferences and context. Index Terms ��� Content search and retrieval, Context model- ing, Ontology, Personalization I. INTRODUCTION HE size and the pace of growth of the world-wide body of available information in digital format (text and audiovis- ual) constitute a permanent challenge for content retrieval technologies. People have instant access to unprecedented inventories of multimedia content world-wide, readily avail- able from their office, their living room, or the palm of their hand. In such environments, users would be helpless without the assistance of powerful searching and browsing tools to find their way through. In environments lacking a strong global organization (such as the open WWW), with decentral- ized content provision, dynamic networks, etc., query-based and browsing technologies often find their limits. Manuscript received October 16, 2006. This research was supported by the European Commission (FP6-001765 ��� aceMedia), and the Spanish Ministry of Science and Education (TIN2005-06885). The expressed content is the view of the authors but not necessarily the view of the aceMedia project as a whole David Vallet, Pablo Castells, and Miriam Fern��ndez are with the Universi- dad Aut��noma de Madrid, Escuela Polit��cnica Superior, 28049 Madrid, Spain (phone: +34-914972284, fax: +34-914972235, e-mail: david.vallet@uam.es, miriam.fernandez@uam.es, pablo.castells@uam.es). Phivos Mylonas and Yannis Avrithis are with the National Technical Uni- versity of Athens, Image, Video & Multimedia Laboratory, 9, Iroon Polytech- niou str., 15773 Zographou, Athens, Greece (email: fmylonas@image.ntua.gr, iavr@image.ntua.gr). Copyright (c) 2006 IEEE. Personal use of this material is permitted. How- ever, permission to use this material for any other purposes must be obtained from the IEEE by sending an email to pubs-permissions@ieee.org. Personalized multimedia content access aims at enhancing the information retrieval (IR) process by complementing ex- plicit user requests with implicit user preferences, to better meet individual user needs [15], [20]. Personalization is being currently envisioned as a major research trend to relieve in- formation overload, since IR usually tends to select the same content for different users on the same query, many of which are barely related to the user���s wish [9]. The combination of long-term and short-term user interests that takes place in a personalized interaction is delicate and must be handled with great care in order to preserve the effectiveness of the global retrieval support system, bringing to bear the differential as- pects of individual users while avoiding distracting them away from their current specific goals. Reliability is indeed a well-known concern in the areas of user modeling and personalization technologies. One impor- tant source of inaccuracy of automatic personalization tech- niques is that they are typically applied out of context. In other words, although users may have stable and recurrent overall preferences, not all of their interests are relevant all the time. Instead, usually only a subset is active at a given situation, and the rest can be considered as ���noise��� preferences. In order to provide effective personalization techniques and develop in- telligent personalization algorithms, it is appropriate to not only consider each user���s queries/searches in an isolated man- ner, but also to take into account the surrounding contextual information available from prior sets of user actions. It is common knowledge that several forms of context exist in the area [23]. This paper is concerned with exploiting se- mantic, ontology-based contextual information, specifically aimed towards its use in personalization for content access and retrieval. Among the possible knowledge representation formalisms, ontologies present a number of advantages [30], as they provide a formal framework for supporting explicit, machine-processable semantics definitions, and facilitate in- ference and derivation of new knowledge based on already existing knowledge. The goal of the research presented herein is to endow per- sonalized mutimedia management systems with the capability to filter and focus their knowledge about user preferences on the semantic context of ongoing user activities, so as to achieve a coherence with the thematic scope of user actions at runtime. We propose a method to build a dynamic representa- tion of the semantic context of ongoing retrieval tasks, which is used to activate different subsets of user interests at runtime, in such a way that out-of-context preferences are discarded. Personalized Content Retrieval in Context Using Ontological Knowledge David Vallet, Pablo Castells, Miriam Fern��ndez, Phivos Mylonas, and Yannis Avrithis T