Browse Prior Art Database

Smart Application Selection Assistant

IP.com Disclosure Number: IPCOM000114910D
Original Publication Date: 1995-Feb-01
Included in the Prior Art Database: 2005-Mar-30
Document File: 4 page(s) / 147K

Publishing Venue

IBM

Related People

Becker, CH: AUTHOR [+4]

Abstract

Disclosed is a process for continuously updating the icons within a file folder according to a simple Artificial Intelligence algorithm running in the background to provide the user at any time with icons representing the applications or program elements that will most likely be needed.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 49% of the total text.

Smart Application Selection Assistant

      Disclosed is a process for continuously updating the icons
within a file folder according to a simple Artificial Intelligence
algorithm running in the background to provide the user at any time
with icons representing the applications or program elements that
will most likely be needed.

      Fig. 1 is an example of the appearance of a Smart Application
Selection Assistant (SASA) window, which is divided into Most Often
Used (MFU), Most Recently Used (MRU), and Most Likely Used (MLU)
subsections.  The MFU subsection contains icons for programs or
program elements most often selected by the user, as determined by a
simple tally program counting the number of times each icon is
selected from anywhere within the operating system.  The MRU
subsection contains icons most recently selected by the user.  The
MLU subsection contains icons most likely to be used next, as
predicted by the artificial intelligence algorithm.  These
subsections may be differentiated by color, font, or background
texture.  The SASA window can be minimized, moved, resized, or
iconized.

      A predictive algorithm to determine the contents of the MLU
subsystem may be developed, for example, using a simple Markov chain
algorithm, with a Markov transition graph consisting of an N by N
array, where N is the number of icons possible for selection.  Any
time an application Ay is selected immediately following an
application Ax, a value at column y in row x is incremented.
Eventually, the relative probability of each of the applications
being chosen immediately after an application Ax is determined by
dividing the total of each element in row x by the sum of all the
elements in row x.  The particular icons to be placed in the MLU
subsystem, and the order in which these elements appear, is chosen to
reflect probabilities calculated in this way (*).

      Extensions of this approach can be used to calculate more
complex relationships.  For example, more recent selections may be
given greater weight than past selections in calculating the
probabilities.  Similarly, the matrix can be extended to allow
predictions based on historical data accumulated before the use of
the currently chosen application.  For example, the program may learn
that the user is more likely to go from an editor to a compiler when
software development steps have been recently performed, and from an
editor to a spell checker when text processing steps have been
recently performed.  Using straightforward and efficient algorithms,
Markov chain approaches can be tuned to make predictions based on
time, history, and usage patterns.

      Fig. 2 is a perspective view of an external selection device
used in an alternate embodiment.  The buttons identified with the
letters F, R, and L are used to select the MFU, MRU, and MLU
subsections, respectively.  While the content of each of these
subsections is determined as described above, the di...