From text to image An experimental multimedia information system based on querying texts in natural language

We see multimedia products of a cultural nature which are destined for the public at large appearing in museums and more generally in the area of Fine Arts, videodisc, CD-ROM and CD-I collections, either as edited products to be commercialised or as works designed to be used in museums. Innovative in the information technology they use, and sometimes also in their aims, these products are time-consuming to build up, because the data that they contain is only brought together for this purpose and, if we want to guarantee the scientific quality of the work, it has to be done by specialists. Would it be possible to build an information system capable of satisfying different types of public, from the same data, coming from different origins, and already built up for another objective? Can data collected by researchers be opened up and accessed by other methods of enquiry than that for which it was initially put together?