Baise, Paul (1994) Cogitator : a parallel, fuzzy, database-driven expert system. Masters thesis, Rhodes University.
The quest to build anthropomorphic machines has led researchers to focus on knowledge and the manipulation thereof. Recently, the expert system was proposed as a solution, working well in small, well understood domains. However these initial attempts highlighted the tedious process associated with building systems to display intelligence, the most notable being the Knowledge Acquisition Bottleneck. Attempts to circumvent this problem have led researchers to propose the use of machine learning databases as a source of knowledge. Attempts to utilise databases as sources of knowledge has led to the development Database-Driven Expert Systems. Furthermore, it has been ascertained that a requisite for intelligent systems is powerful computation. In response to these problems and proposals, a new type of database-driven expert system, Cogitator is proposed. It is shown to circumvent the Knowledge Acquisition Bottleneck and posess many other advantages over both traditional expert systems and connectionist systems, whilst having non-serious disadvantages.
|Item Type:||Thesis (Masters)|
|Uncontrolled Keywords:||Anthropomorphic machines, Knowledge, Intelligence, Manipulation, Expert system, Knowledge acquisition bottleneck, Machine learning, Database, Computer, Cogitator|
|Subjects:||Q Science > QA Mathematics > QA75 Electronic computers. Computer science|
Q Science > QA Mathematics > QA76 Computer software
|Divisions:||Faculty > Faculty of Science > Computer Science|
|Supervisors:||Clayton, Peter and Murali, Venkat|
|Deposited By:||Philip Clarke|
|Deposited On:||17 Oct 2012 13:14|
|Last Modified:||17 Oct 2012 13:14|
0 full-text download(s) in the past 12 months
Repository Staff Only: item control page