Abstract:
Response time in an infonnation system can be improved by reducing the number of blocks
accessed when retrieving a document set. One approach is to restructure the document base in
such a way that similar documents are placed close together in the file space. This ensures a
greater probability that documents will be co-located within the same block. This thesis is based on a file sequencmg algorithm proposed by Lowden [Lowden 1985) and
attempts to 1l'lodiiY and improve the algorithm proposed. A series of three sequencing algorithms
are considered. The effect of associating weights to key tenns during document sequencing is the
major area .of interest. Simulation with computer programs and statistical theories are used to
analyze the performance of these algorithms.
