Compactly encoding unstructured inputs with differential compression (2002) [24 citations — 8 self]
|
Popular Tags
No tags have been applied to this document.
BibTeX | Add To MetaCart
@ARTICLE{Ajtai02compactlyencoding,author = {Miklos Ajtai and Randal Burns and Ronald Fagin and Darrell D. E. Long and Larry Stockmeyer},
title = {Compactly encoding unstructured inputs with differential compression},
journal = {Journal of the ACM},
year = {2002},
volume = {49},
pages = {318--367}
}
Years of Citing Articles
Abstract:
The subject of this paper is differential compression, the algorithmic task of finding common strings between versions of data and using the commonality between versions to encode one version compactly by describing it as a set of changes from its companion. A main goal of the work is to present new differencing algorithms that operate at a fine granularity (the atomic unit of change), make no assumptions about the format or alignment of input data, and use linear time and give good compression on version data that typically arises in practice. First we review existing differencing algorithms that compress optimally but use time proportional to n 2 in the worst case and space proportional to n in the best and worst case. Then we present new algorithms, which do not always compress optimally but use considerably less time and space than existing algorithms. One new algorithm runs in O(n) time and O(1) space in the worst case (where each unit of space contains dlog ne bits, and the space to hold the input data is not included). We introduce two new techniques for differential compression, and we apply these to give additional algorithms that improve the compression performance of the linear-time algorithm and the time performance of the quadratic-time algorithm. Having presented these algorithms, we experimentally explore their properties, such as time and compression performance, by running them on actual versioned data. In these experiments, the new algorithms run in linear time and constant space, and their compression performance closely approximates that of previous algorithms that use considerably more time and space. Finally, we present theoretical results that limit the compression power of differencing algorithms that are restricted to making only a single pass over the data.
