Home | Business News | Browse by Publication | C | Computer Technology Review

Interactive exploration of non-indexed data.

Article Excerpt
Creating and storing large archives of rich media such as digital photographs, videos, X-rays, MRIs and CAT scans is becoming easier: for example, a digitized medical institution generates approximately 12 terabytes of medical images annually. However, finding a vaguely-specified object in such an archive is very difficult. Unlike text data, which is easily indexed by search engines, rich media is difficult to search by content. In such cases, one has no alternative but to perform a brute-force search--examining each item in the repository to determine whether it matches the user's needs.

Unfortunately, native implementations of brute-force search are very inefficient for several reasons. First, testing the query against every item in a large collection can be prohibitively expensive in terms of computation. Second, the available network bandwidth may be insufficient to ship all of the data to the user's computer, particularly when the data repository is distributed at several sites on the internet. Third, the effort involved in shipping and processing most of the data is wasted since only a very small fraction of the data is likely to be relevant to any given query. For these reasons, most institutions simply store the data without offering means to search it. This problem becomes worse as...

View this article FREE - Now for a Limited Time, try Goliath Business News
Free for 3 Days!