Abstract: The present invention is embodied in a method and system for compressing data using clustering techniques. The invention provides compression of a data ensemble composed of individual data sets. Each individual data set has associated data set context and history contexts. The present invention groups similar data set contexts and history contexts together so as to achieve a reduced total encoding cost. This total cost is minimized by selecting the clustering of the data set and history contexts such that the models or codebooks used to compress tokens are shared between tokens assigned to the same context clusters. Using clustering techniques, the present invention alternately varies the data set context clustering and history context clustering until a minimum total cost is achieved.

Abstract: A system and a method for generating a find list of target-entries that can be searched based on a received search-entry. The techniques used to generate and search the find list allow meaningful results to be obtained even if the search-entry contains imperfections. Target-entries are selected from the find list based on the initial characters of the search-entry. The target-entries are then compared to the search-entry using a variety of scoring heuristics. Each scoring-heuristic assigns a value based on the results of the comparison. The value proportionately indicates the closeness of the match between the target-entry and the search entry. The value obtained from each scoring-heuristic is summed to identify the score for each target-entry. The selected target-entries are then ordered based on the scores.