Combining Multimedia Retrieval and Text Retrieval to Search Structured Documents in Digital Libraries

Digital Libraries usually contain a large collection of structured multimedia documents. At present text may be dominant in many applications, however the relevance of other media types such as image, audio and video increases steadily. An important functionality of a digital library in this respect is the retrieval of relevant multimedia documents or relevant parts of multimedia documents. To this end, an efficient combination of automatic text retrieval, retrieval in meta data (usually created manually) and content based retrieval on multimedia data is needed. In this position paper, we will argue, that a sophisticated query language for an (at least structurally) object-oriented database is a suitable basis for an application specific user-interface built on top of it. We will sketch the requirements for such a query language and present our research approach.