Embodiments of the present disclosure set forth methods for selecting a preferred data set. The methods include generating a joined relation based on a first relation having a first join attribute and a first existence probability attribute, and a second relation having a second join attribute compatible with the first join attribute and a second existence probability attribute, wherein the joined relation comprises a skyline probability attribute based at least in part on the product of a second value of the first existence probability attribute and a third value of the second existence probability attribute; and selecting, by one or more processors, the preferred data set from the joined relation based on a comparison of the first value of the skyline probability attribute and a predetermined threshold.
展开▼
