With the dramatic increase of knowledge management systems in commercial applications, many knowledge management systems are now becoming legacy systems. These knowledge management systems contain knowledge depositories that are often large and unstructured. Therefore maintaining knowledge management systems is a very difficult task. This study introduces a Hopfield neural net-based software tool, which can be used for facilitating maintenance of a knowledge depository in knowledge management systems. This neural net software will automatically categorise knowledge pieces of a knowledge depository according to its lexical similarity. Clusters will be made as an output of this neural net software. Having clusters out of a knowledge depository will facilitate the knowledge management system maintainer since categorised knowledge pieces are easier to understand than disordered knowledge pieces of a knowledge management system.
[1]
David Sharon.
Meeting the Challenge of Software Maintenance
,
1996,
IEEE Softw..
[2]
G. A. Miller.
THE PSYCHOLOGICAL REVIEW THE MAGICAL NUMBER SEVEN, PLUS OR MINUS TWO: SOME LIMITS ON OUR CAPACITY FOR PROCESSING INFORMATION 1
,
1956
.
[3]
Niv Ahituv,et al.
Principles of information systems for management
,
1982
.
[4]
Anneliese Amschler Andrews,et al.
Program Comprehension During Software Maintenance and Evolution
,
1995,
Computer.
[5]
Laurene V. Fausett,et al.
Fundamentals Of Neural Networks
,
1994
.