SOURCE:

Faculty: Engineering
Department: Electronic And Computer Engineering

CONTRIBUTORS:

Nwobodo, L.O;
Inyiama, H.O;

ABSTRACT:

A Dynamic Expert System (DES) is an Expert system whose knowledge base is dynamically updated with time in order to render its expert advice relevant at all times. In this Dynamic Expert System (DES), experienced staff among the workforce constitutes the domain expert, the researcher functions as knowledge engineer, who elicits knowledge from the domain experts. The elicited knowledge is further verified and validated by the workforce peers namely the principal engineer and the director. The validated knowledge is rated by the peers (grade 1, grade 2, grade 3) and then codified and stored in the knowledge repository by the knowledge engineer. Quarterly, the workforce is exposed to professional update courses towards enhancing their knowhow. Where funds are available, research and development efforts are also undertaken. The knowledge elicitation exercise is repeated every quarter in search of newly acquired knowhow. The knowledge base is then modified as appropriate and made to be continuously up to date. The DES was subjected to Monte Carlos simulation to see how it would fare for the next five years (20 quarters). A remuneration package for new IP creation was crafted by the researcher such that the remuneration paid to the workforce is always less than the contribution of the new IPs to the company’s bottom line. This enables the company to acquire new IP at a reduced cost than its market worth. A linear regression model of the new IP was found and the coefficient of determination was 98.63% meaning that the model can be used for forecasting new IP. Over the 5years of simulation, a total of 282 new IPs were discovered causing the company a remuneration package of N808,109.50. However, the expert valuation of those new IPs gave their market worth to be N7,680,300. This means that the company produce the new IPs cheaper than their market price by as much as N6,872,190 within the 5years of simulation. The software used for the Dynamic Expert System (DES) inference Engine is written in Java Programming Language. DES was validated by comparism with other related projects and it was found that DES performed better than others in five (5) of the ten (10) areas of comparism used. Faculty of Engineering research output was used to model patentability of research findings. 100 (one hundred) randomly selected research output of the faculty was evaluated for patentability, by expert valuers in terms of novelty, contribution to knowledge, ease of use and cost of adoption. It was found that only 10% out of the 100 completed projects could be patented. Patentability was then modeled using multiple regression analysis. It was found that the higher the number of good quality projects completed annually within the faculty of Engineering, the higher the proportion of them that would be patentable and the higher the money accruable to the faculty from research work. The patentability model can be used in What-if analysis to determine research output earnings in the Faculty of Engineering as their quality and patentability increases. It is highly recommended that the faculty of Engineering in any university should insist on high quality (Patentable) timely research output because it greatly enhances their earning capacity.