MULTIDIMENSIONAL DATA OBJECT SEARCHING USING BIT VECTOR INDICES

摘要

Before searching the multidimensional feature space where data object is thrown, each dimension in the feature space is divided into several sections. When kueri is received, the single section where kueri is piled up, is selected from each dimension. Next, data object gathering which includes just the data object which is piled up to the section which is selected is reduced is selected. Next, the whole gathering of data object in the feature space it is not, this data object gathering which is reduced is used, the match for kueri is decided.