Subspace similarity search: efficient k-NN queries in arbitrary subspaces

摘要: There are abundant scenarios for applications of similarity search in databases where the objects is defined a subset attributes, i.e., subspace, only. While much research has been done efficient support single column queries or full space, scarcely any subspaces provided so far. The three existing approaches variations sequential scan. Here, we propose first index-based solution to subspace arbitrary subspaces.