Browse Prior Art Database

Almost Equal Search Qualification and Error Recovery

IP.com Disclosure Number: IPCOM000044570D
Original Publication Date: 1984-Dec-01
Included in the Prior Art Database: 2005-Feb-06
Document File: 1 page(s) / 13K

Publishing Venue

IBM

Related People

Convis, DB: AUTHOR [+3]

Abstract

This technique provides a tool to assist the user of a data processing system when an input string of characters would not otherwise be recognized by the system as a valid procedure or name of a document, file, etc. Current computer systems function efficiently when used with precision. Each user input is matched against the possibilities, and when recognized, the associated action is taken. Without an equal match of input characters to a valid possibility, the system returns a failure message. Failures because of miskeying, misspelling, not remembering exactly, not being told exactly, or not knowing the correct entry are not uncommon.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 1

Almost Equal Search Qualification and Error Recovery

This technique provides a tool to assist the user of a data processing system when an input string of characters would not otherwise be recognized by the system as a valid procedure or name of a document, file, etc. Current computer systems function efficiently when used with precision. Each user input is matched against the possibilities, and when recognized, the associated action is taken. Without an equal match of input characters to a valid possibility, the system returns a failure message. Failures because of miskeying, misspelling, not remembering exactly, not being told exactly, or not knowing the correct entry are not uncommon. This article provides an efficient technique for comparing an input string from the user to a data string from a list of possibilities to determine if the data string is almost equal to the input string. This technique combines the previously published method [*] with a dual thresholding technique to identify almost equal candidates. The present technique is given an input string and length, and the data string and length. An indicator is returned set to "almost equal" if the match rate between the strings is greater than or equal to the calculated threshold; otherwise, the indicator is set to "not almost equal". This threshold is set as a percentage of the longer string (input or data) truncated to an integer. The method of [*] is used to determine the match rate between the input and data strings. A length-dependent threshold is used to identify the match rate required to accept the data string as almost equal. THRES = 3/4 (Max(DLEN,ILEN)) where: DLEN - Data string length ILEN - Input string length THRES - Threshold value (truncated to an integer) This dual thresholding technique incorporate...