Incremental Acquisition of Conceptual Indices for Multimedia Design Documentation

Information retrieval systems based on conceptual indexing can access the underlying meaning of text, graphics or videotaped documents. Since conceptual indices represent the semantics of a piece of information, it is difficult to extract them automatically from a document, and it is tedious to build them manually. We present a method to acquire and refine conceptual indices in the context of Dedal, a system that facilitates the indexing and retrieval of text, graphics and videotaped design documents in the mechanical engineering domain. Our approach is to use an underlying model of the domain covered by the documents to constrain the user's queries. This facilitates question-based acquisition of conceptual indices: converting user queries into indices which accurately model the content of the documents, and can be reused. We demonstrate the relevance and coverage of the acquired indices through experimentation.