A multilevel signature approach for retrieval of unformed data

A multilevel signature approach for retrieval of documents from text or unformatted databases is presented. This approach provides a solution to the false drop problem by eliminating the nonqualifying documents at several levels. Each level acts as a filter which provides input to the next level. A prototype system has been developed to test the effectiveness of this approach. Each search process was separately implemented and tested over a database of 50 documents of 5 kB per document. The main purpose of these experiments was to measure the number of false drops in the first two levels. This was done by comparing the results of the first two processes to the third-level process which should result in no false drops. The experiments on the selected queries have so far shown no false drops