Clustering records in information retrieval systems. (c1995)

Abstract

This project is a study of the space density of a file and how it affects retrieval time. The number of blocks that are retrieved when a query is made to the document is used as a measure of response time. The space density is an indication of how close related records are placed. The hannning distance of a file is used as a measure of space density. A sequencing algorithm based on threshold values that reorders records in a file so as to increase average record similarity is used to obtain varying space densities. Simulation experiments conducted proved that a great reduction in response time is yielded after the restructuring of a file with a reasonable amount of work required in sequencing. Other terms such as block size and terms in a query which affect response time are studied. Results from the experiments are shown graphically. Statistical methods are used to confirm the experimental results.