| This thesis explains and compares the following three granule-based computing techniques: Granule Based Appriori, Granule Based Appriori Lattice, and Granule Based Lattice. The latter two algorithms are new and are proposed in this paper. This thesis also explains a technique to store the data in a granule form in a data warehouse.; A granule is a compact list of tuple names having the same attribute value and is represented in binary form. With the data converted to granules, the relationships are determined by the intersection of the granules. The operations performed on the granules are natural instructions for the computer. This has a great significance in the speed of computation. It is costly and time consuming to access and process the data to find patterns. Thus, it is important for the storage system to not only provide fast response time but also provide the data in a format most suitable for processing. |
