Combining Heterogeneous Lexical Resources

One of the main tasks of the Natural Language Processing Group at the Faculty of Mathematics, University of Belgrade is the development of various lexical resources. Among them the two most important ones are: the system of morphological dictionaries of Serbian (SMD) in Intex format and the Serbian wordnet (SWN) developed in the scope of the Balkanet project. Although these two resources represent dictionaries of a different type, developed using different models, each of them contains information that can either be incorporated in to the other dictionary or that can be used in its development. In this paper we will outline some of the most interesting examples. W e also present an integrated programming tool that enables the integration of these diverse lexical resources, as well as possible applications. We envisage the use of these resources in defining and linking lexical data in a way that will enable their more effective retrieval, integration, and reuse across various Web applications.