Vertical compression algorithms for sequentially processed statistical files
Horizontal data compression eliminates redundancies or regularities that occur within individual records. Suppression of trailing blanks and leading zeros are examples. Vertical compression eliminates regularities that occur across consecutively stored records. Prefix and suffix compression are examples. Two new vertical compression algorithms, VRE and HVRE, are presented in this paper. They are based on a combination of character matrix transposition (where rows are initially identified with records) and horizontal compression algorithms (run-length encoding and Huffman encoding). Experimental and theoretical results are presented that show the performance of VRE and HVRE is superior to that of some reputable commercial algorithms. Specifically, these are the compression algorithms of the ADABAS, IDMS and SHRINK/2 data management systems. VRE and HVRE are best suited for compressing statistical files which are sequentially processed and batch updated. They may also be used for file archival and for compressing randomly processed files as well.
- Research Organization:
- Florida Univ., Gainesville (USA). Dept. of Computer and Information Sciences
- DOE Contract Number:
- AS05-81ER10977
- OSTI ID:
- 6663671
- Report Number(s):
- DOE/ER/10977-T4; ON: DE84015040
- Resource Relation:
- Other Information: Microfiche only, copy does not permit paper copy reproduction
- Country of Publication:
- United States
- Language:
- English
Similar Records
Proposed first-generation WSQ bit allocation procedure
GOTTCHA Database, Version 1