Abstract: Systems and methods are provided for compressing data structures. One exemplary method includes accessing a target data structure defining multiple columns, and filtering the columns based on a cardinality of terms in each of the columns. The method also includes, for each filtered column, sorting the data structure by the column, compressing the sorted data structure, and identifying the filtered column as a candidate column when the size of the compressed and sorted data structure is less than a baseline size. The method further includes, for each pair of candidate columns, sorting the data structure by the pair of candidate columns, compressing the pair-sorted data structure, and designating the compressed pair-sorted data structure as an object data structure and the pair of candidate columns as a sorting column pair, when said compressed pair-sorted data structure includes a smallest size compared to sizes of other compressed pair-sorted data structures.

Abstract: Systems and methods are provided for compressing data structures. One exemplary method includes accessing a target data structure defining multiple columns, and filtering the columns based on a cardinality of terms in each of the columns. The method also includes, for each filtered column, sorting the data structure by the column, compressing the sorted data structure, and identifying the filtered column as a candidate column when the size of the compressed and sorted data structure is less than a baseline size. The method further includes, for each pair of candidate columns, sorting the data structure by the pair of candidate columns, compressing the pair-sorted data structure, and designating the compressed pair-sorted data structure as an object data structure and the pair of candidate columns as a sorting column pair, when said compressed pair-sorted data structure includes a smallest size compared to sizes of other compressed pair-sorted data structures.