After all the raw information is available, the requirements engineer or
elicitor must begin to identify the abstractions present in this
information. These abstractions can be used as the basis for a variety of
requirements documents, including chapters of the specification, classes in
an object-oriented model of the requirements, node names in a hypermedium
representation of the requirements, etc.

<p>
Recently, Leah Goldin, one of Berry's Ph.D. students developed
AbstFinder, a tool that assists the requirements engineer in finding
abstractions in the confusing mass of documents the client delivers to
the requirements engineers at the start of the development of a
software-based computing system. Her thesis described a method for
using the tool and presented an industrial strength case-study that
demonstrated the effectiveness of the method and the tool for the task
of abstraction identification.

<p>
It is now time to build a production quality version of the tool, with a
groovy WIMP interface. Any takers?

<p>
For further information:

<p>
Goldin, L. ``A Method for Aiding Requirements Analysts in Requirements
Elicitation for Large Software Systems'', Ph.D. Dissertation, Faculty
of Computer Science, Technion, Haifa, Israel, May, 1994
&nbsp;
<a href =
"ftp:FTP_SITE/students.theses/goldin.thesis/"> FTP
</a>

<p>
Leah Goldin's Ph.D Thesis and <tt>AbstFinder</tt> Source Code
&nbsp;
<a href =
"ftp:FTP_SITE/software.distribution/goldin.thesis/"> FTP
</a>

<p>
Goldin, L. and Berry, D.M., ``<tt>AbstFinder</tt>, A Prototype
Natural Language Text Abstraction Finder for Use in Requirements
Elicitation'' <cite>Automated Software Engineering</cite>, 4:4, 375-412,
October, 1997
<a href =
"FTP_SITE/reprints.journals.conferences/abstfinder.journal.paper.pdf"> PDF
</a>
&nbsp;
<a href =
"FTP_SITE/reprints.journals.conferences/abstfinder.journal.paper.ps.Z">
ps.Z
</a>