Browse Prior Art Database

Categorization via the High Performance File System

IP.com Disclosure Number: IPCOM000110098D
Original Publication Date: 1992-Oct-01
Included in the Prior Art Database: 2005-Mar-25
Document File: 2 page(s) / 51K

Publishing Venue

IBM

Related People

Davis, A: AUTHOR [+3]

Abstract

This invention provides for the use of a hierarchically organized file system's optimized file search capabilities to be used to support searches based on categories and keywords.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Categorization via the High Performance File System

       This invention provides for the use of a hierarchically
organized file system's optimized file search capabilities to be used
to support searches based on categories and keywords.

      The figure depicts the parallel between categories and
keywords, and subdirectories and files that are utilized in this
solution.  As can be seen in the figure, the directory is named
according to the category and the files within the directory are
named according to the keywords.  A category can contain other
categories (subdirectories).  A file contains whatever information
about the components being "tagged" by the keyword, so that the
components can be retrieved based upon the keyword being used in a
search expression.

      Since the High Performance File System (HPFS) supports long
names, essentially any category and keyword name can be supported.
Also, since the HPFS has been optimized for quick file retrieval,
getting to the right component information is quick.

      As an example, to find objects tagged with the keyword
"BiDirectional", within the category "NLS", the file with the path
d:\NLS\ BiDirectonal would be accessed.

      Any file system, and not just the HPFS file system, that
supports long names and subdirectories can be used in this way.

      Categorization information can be distributed across a LAN or
WAN, allowing essentially unlimited amounts of information to be
accessed.