Future prospects for network‐based multimedia information retrieval

This paper asks the reader to consider future information retrieval services that return multimedia documents containing any combination of text, graphics, still raster images, audio, and motion video in response to user queries. It then argues that the emergence of generally available multimedia information services depends on four things: (1) close to ubiquitous high‐bandwidth networks; (2) inexpensive user appliances capable of handling multimedia; (3) adoption of standards for representation, compression, packaging and transport of multimedia information; and (4) development of a corpus of multimedia information and associated infrastructure for organizing and searching it. After some explanation, it asserts that the first three are already happening and expresses reserved optimism about the fourth.