Browse Prior Art Database

Storage and Retrieval of Very Large Tables Used for Sorting Checks

IP.com Disclosure Number: IPCOM000118147D
Original Publication Date: 1996-Oct-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 4 page(s) / 156K

Publishing Venue

IBM

Related People

Stadler, FL: AUTHOR

Abstract

Disclosed is a method for storing very large tables of data used to sort checks and also for retrieving that data within the time constraints of check sorting devices. The method allows banks and other financial institutions to use very large data files, such as their complete account number list or the full list of bank numbers, in sorting Magnetic Ink Character Recognition (MICR) encoded items.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 34% of the total text.

Storage and Retrieval of Very Large Tables Used for Sorting Checks

      Disclosed is a method for storing very large tables of data
used to sort checks and also for retrieving that data within the time
constraints of check sorting devices.  The method allows banks and
other financial institutions to use very large data files, such as
their complete account number list or the full list of bank numbers,
in sorting Magnetic Ink Character Recognition (MICR) encoded items.

      This method allows compression of very large tables of numbers
(on the order of 1 million digits to well over 100 million digits)
such that they can be stored within the direct memory of check
sorting equipment and accessed with the available pocket selection
timing window.  The technique typically results in compression ratios
of one to  two orders of magnitude and retrieval times of a few
milliseconds on current check sorting equipment (IBM 3890-XP
sorters).

      The method is comprised of two separate phases: a compression
phase and a retrieval phase.  The compression phase is typically
performed on a large, mainframe system (although it could also be
performed on a personal computer).  The retrieval phase is performed
within the processor of the sorter itself.  Retrieval can also be
done by mainframe applications, if desired for compatibility
purposes.

      The input to the compression phase is a table of numbers and
optionally a set of values associated with each number.  The numbers
can be account numbers, bank routing-transit numbers, or any similar
data.  In some applications, the purpose of the table is simply to
identify if a specific number is contained in the table (for example,
if a number is a valid account number for the bank).  In this case
there is no associated data.  In other cases, some additional data
(such as pocket code, float value, etc.) may be associated with each
number.  This method works with either form of table.

      The first step in the compression phase is to sort the numbers
into sequence (either ascending or descending; the example in this
disclosure assumes ascending sequence).

      The second step is to calculate the arithmetic differences
between the successive numbers in the sorted table.  These
differences are calculated by subtracting from each number the value
of the prior number.  The differences are never larger than the
numbers themselves and because of the patterns generally used in
assignment of account numbers, the differences are well suited to
further compression with industry standard data compression
techniques.

      The third step is to segment the resulting table of differences
into multiple, smaller tables.  The count of differences to be
included in each segment is determined by the speed of the processor
that will be used for retrieval of the numbers.  Slower processors
require smaller table segments and faster processors can handle
larger segments.  Each tabl...