Quote: first use of training for lossless compression of tables; 100:1 compression; determine dependency tree of columns; group columns that compress well

Quotation Skeleton

We study the problem of compressing massive tables, … can be put into production use and achieve … GB. We devise a novel compression strategy--training for … The system we built, pzip … [achieves] a two-fold improvement over existing solutions [in time as well as space). .... [p. 176] The idea is to construct a compression plan for the … We further assume and exploit dependencies in the … compress well together; and (2) explicitly, by determining … ours is the first known instance of applying … [p. 177] Differential dependency is an explicit dependency between … [A column is differentially dependent on another column if the difference between columns compresses better than the column itself] …   Google-1   Google-2

Copyright clearance needed for quotation.

Related Topics

Topic: compressed data (16 items)
Topic: data compression algorithms (53 items)
Topic: database implementation (18 items)