Multidimensional information arrangement: a hybrid approach of tree and keyword based classifications

摘要

Classification and arrangement of digital information (entity) are essential to encourage entity sharing. We propose multidimensional arrangement for this purpose. Corresponding to an independent concept, a balanced tree is organized. Entities are classified using multiple trees. Overview of the stored entities becomes possible and the retrieval cost is reduced as well as the management cost. The retrieval flexibility also becomes high since number of search paths are available to the same set of entities. We develop a cost model for the proposed method and perform evaluation through experiment.