Abstract:
The general objective of this study is to construct predicting model and integrate automatically with the knowledge base system, in order to increase the efficiency in the diagnosis and treatment of common cattle disease. Different researches, articles, journal papers and guidelines that have been done on cattle disease diagnosis and treatment, data mining, KBS, knowledge acquisition for KBS and integration of data mining results with KBS were reviewed. Besides of this, related works were reviewed to identify the gap and formulate research questions of the study. Mixed (quantitative and qualitative) research design was used, integrated (manual and automated) knowledge acquisition techniques were used to acquire knowledge, rule based knowledge representation approach was used to represent knowledge in the knowledge base.
Away from each researchers and as to the review of the researcher, no study was done at the area throughout the Amhara Region animal health center using data mining techniques. For this study, the researcher see that the limitation of the existing system is not considering analysis of data in detail.
In this study, rule based diagnosis and treatment of cattle disease knowledge based system is proposed. The system is aimed at utilizing hidden knowledge extracted by employing an induction algorithm of data mining, specifically JRip from the sampled BAHC dataset. The Integrator application, then links the model created by JRip classifier to knowledge based system so as to add knowledge automatically. In doing so, the Integrator understands the syntax of JRip classifier and Prolog and converts from rule representation in JRip to Prolog understandable format.
To do this, java programming was used to integrate WEKA result with the Knowledge Based System automatically. And also ‘swiweka’ is used as an interface that allows the use of WEKA API for classification; weka.jar, Weka _src.jar are used to construct a model when called from interface through swiweka package, JPL library to connect the Java layer with the Prolog lay
Finally, as a result, the proposed system can perform in the absence of domain experts with 98.7% accuracy of recall evaluation results which indicates that the KBS has the necessary knowledge for diagnosis and treatment of cattle disease which in turn implies that the study was effective in acquiring knowledge. Besides of this, the proposed system achieves 85.8% of the users’ acceptance which in turn implies that the proposed system could be operational if it could be implemented.
