A Hierarchical Approach to Object Comparison Axel Hallez and Guy De Tr´ e Ghent university, Dept. of Telecommunications and Information Processing St.-Pietersnieuwstraat 41 9000 Gent, Belgium Axel.Hallez@UGent.be, Guy.Detre@UGent.be Abstract. Comparing objects is a very general problem domain. The way objects are compared can diﬀer a lot, in the way objects are related to each other as well as in the meaning that is attributed to the result of a comparison. In this paper we develop a theoretical framework for com- paring objects. This framework can be tailored to the speciﬁc needs of an application through the choice of appropriate operators and evaluation domain (i.e. the domain in which the comparison results are expressed). 1 Introduction Object comparison is not a trivial problem by any means; several problem do- mains that are heavily researched topics on their own have to be combined in order to construct a comparison scheme: construction of similarity functions for individual properties, construction of aggregation operators, dealing with pref- erences, compensation, etc... Most of these topics have been researched quite extensively now and although new approaches are proposed on a regular basis, many solutions are readily available for use in applications. However, one of the biggest problems that remains, one that cannot be automated, is to ﬁnd out which techniques are best suited for comparing objects in a particular situation, eg. which aggregation operators to use, ﬁnding out the suitable weights for the partial results, etc... In order to do that one has to gain a deep understanding of both the application domain and the techniques that are available for comparing objects. The usual practice is to ﬁnd a domain expert and a technical expert and let them work together to ﬁnd out how objects compare to each other. It would of course be much more convenient if the domain expert would be able to work with no or less help of a technical expert. Therefor it would be handy to have a (software) tool at hand that allows a domain expert to construct a com- parison scheme and evaluate it. The easier this process gets, the more feasible it becomes to test diﬀerent comparison methods and to compare the results. So we want to build a framework for the construction of comparison schema where a developer (i.e. the domain expert) can easily specify how objects are to be compared, without having to implement everything from scratch. In this paper we present a theoretical foundation of a generic framework for the construction of comparison schema. The framework is generic in several respects. Firstly, the framework can be used for a wide range of comparison P. Melin et al. (Eds.): IFSA 2007, LNAI 4529, pp. 191–198, 2007. c  Springer-Verlag Berlin Heidelberg 2007