Browse Prior Art Database

Writeable Adaptive Transient Dictionary

IP.com Disclosure Number: IPCOM000049657D
Original Publication Date: 1982-Jul-01
Included in the Prior Art Database: 2005-Feb-09
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Glickman, D: AUTHOR [+2]

Abstract

In a spelling verification environment, the main dictionary is of a finite size; that is to say, the dictionary does not contain every word in the language nor does it contain specialized words or persons' names unique to any specific industry or company. The user may thus encounter a high "false alarm" rate where specialized words and names are flagged as potential spelling errors. The purpose of the Writeable Adaptive Transient Dictionary (WATD) is to allow the user to add words, unique to the user's environment, into the system to be used for word verification, thus avoiding "false alarms".

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 2

Writeable Adaptive Transient Dictionary

In a spelling verification environment, the main dictionary is of a finite size; that is to say, the dictionary does not contain every word in the language nor does it contain specialized words or persons' names unique to any specific industry or company. The user may thus encounter a high "false alarm" rate where specialized words and names are flagged as potential spelling errors. The purpose of the Writeable Adaptive Transient Dictionary (WATD) is to allow the user to add words, unique to the user's environment, into the system to be used for word verification, thus avoiding "false alarms".

A WATD of 500 provides acceptable storage/performance characteristics. Because the WATD is used for user additions "on the fly", it must reside in the system random-access memory rather than diskette storage. The WATD entries are encoded and stored in ascending sort sequence.

If the user continually enters new words into the WATD, eventually all of the available slots will be exhausted. If a new word is then entered, the oldest entry in the WATD will be deleted to make room for this new word. The method to discriminate between the entries in the WATD to determine which entry is chronologically the oldest is through the use of a data field (8 bits long) appended to each word in the WATD. This data field is called the Event Counter (EC). The EC is event driven and is independent of chronological time. The EC is updated by new words being added to the WATD (events).

A simplified structure of the WATD is shown below:

A word 1 Event Counter 1 Event Byte

B word 2 EC2

C word 3 EC3

D word 4 EC4

E word 5 EC5

The Event Byte (only one per WATD) is initially set to zero.

Each time a new entry is to be added to the WATD. The EC associated with that new entry is set to a quantity called the EC-high value
(H). The Event Byte is then incremented by one. The Event Byte is interrogated to determine if all the WATD entries excluding the new entry should have their EC values decremented by one. By decrementing the EC values, inspection of each value can chronologically relate its age with respect to the other entries. If no free space is available in the WATD, the oldest entry is deleted and the new entry is sort sequence inserted into the WATD. The method in which the event counter is interrogated is to compare the Event Byte value against a quantity...