Knowledge Based System for Ground Water Potential Prediction

Abstract

Groundwater potential prediction here refers to the total amount of permanent storage water that exists in the aquifers. Groundwater potential is the function of the porosities of the rocks and amount of open space in rocks that could store water. Most of Ethiopia population lives in rural area have not got sufficient water for drinking, irrigation and livestock raring. Lack of pure water in the rural area and town is the major problem for industrial development and economically. Using ground water is the solution for the problem of lack addressing sufficient water for people in Ethiopia. The aim of this study was developing knowledge base system for groundwater potential prediction. To do so, a design science research methodology was used. Development of KBS for GWPP to improve the quality of decision making for fresh hydrologist and geologist to predict groundwater potential effective and efficiently in the shortage of domain expert. To achieve this objective knowledge was acquired by using sample size from Jimma University Institute of Technology, Faculty of Civil and Environmental Engineering, Jimma Zone and Jimma Town water and Energy organizations. Purposive sampling techniques used to selected domain expert and secondary was collected from difference of journal article, groundwater directive, manuals, books and different website working on assessment of groundwater potential.KBS is developed by using conceptual model and representation which is supported by decision tree easily to understand and interpret the steps of groundwater potential prediction. Prototype is developed based on the conceptual model using WINprolog version 6.4.0. production rule if then rules,and forward chaining reasoning mechanism to inference engine rule and appropriate decision making. Performance of the prototype KBS have got excellent acceptance by system evaluator and 85.7% users satisfied by the performance of the prototype developed. Finding of this research was performance of system evaluated by using confusion matrix predict validation techniques by using twenty two parameters of predicting groundwater potential then result of validation parameters of prototype is 84.33% accurate to predict GWPP. Future work proposed by researcher predicting spring water potential continuity by using knowledge based system