The amount of information available on the Internet, primarily by way of the World Wide Web, is truly staggering. According to one measurement, in February 1999 there were about 800 million web pages publicly available on about 3 million web servers, for a total of approximately 9 terabytes of data (Lawrence, S. and Giles, S. L. 1999). These enormously large numbers are a testament to the success of the Internet in providing a way for people all around the world to share information and communicate with each other.
In addition, since the study does not even mention the existence of other forms of data available, such as audio or video (movie) files, one could infer that the amounts of data publicly available in other multimedia forms is overshadowed by the amount of text and image data. Clearly, text is the dominant format of data on the web.
The main reason for this propensity towards text was because of the capabilities of the technology available. In the past, network bandwidth was generally low, as was disk space, memory space, and processing power; for this reason, non-textual data such as images or sounds could not be supported by most computing platforms. According to Besser, by today's standards, storage capacity was miniscule, networks were unbearably slow, and visual display devices were poor. Recent increases in storage capacity, network bandwidth, processing power, and display resolution have enabled a tremendous growth in image database development.
With computer technology improving at a phenomenal pace, the technology limitations which dictated the predominant use of text on the Internet in the past are lessening. In the very near future, non-textual data will be as common a format for publicly available data as text is now.
In light of these trends, it is important to review the state of the art of the retrieval of such non textual, multimedia data. Text information retrieval is already well established; most data retrieval systems, such as web search engines, are text retrieval systems. However, multimedia information retrieval is less established. There are a number of open issues involved in such retrieval.
What are Multimedia Retrieval systems
It involves retrieval of full motion video, but it is sometimes refer to in the context of the retrieval