One　of　the　main　obstacles　facing　current　data　mining　techniques　is　attribute　reduction.　This　paper　discusses　the　basic　concepts　of　rough　set,　and　studies　two　rough　approximations　operators　under　neighborhood　systems.　An　attribute　reduction　method　based　on　rough　set　theory　and　neighborhood　systems　is　presented.　The　experimental　results　show　that　the　method　of　attributes　reduction　with　rough　sets　and　neighborhood　system　is　feasible　and　valid.