Pest Activity Prognosis in Rice Fields Using Fuzzy Expert System Approach

• Main Author: Nureize, Arbaiy
• Format: Thesis
• Language: English; English
• Published: 2004
• Subjects: QA76.76 Fuzzy System.
• Online Access: https://etd.uum.edu.my/1398/1/NUREIZE_ARBAIY.pdf; https://etd.uum.edu.my/1398/2/1.NUREIZE_ARBAIY.pdf; https://etd.uum.edu.my/1398/

Fuzzy Logic (FL) is a form of knowledge representation which is appropriate for notions that cannot be define precisely, but depends upon its context. An Expert System (ES) is a computer program that uses human knowledge to solve problems in typical tasks, which normally requires human intelligence. As knowledge involved in pest management is imperfect, vague and not completely reliable, fuzzy logic is integrated in this expert system to deal with the approximate reasoning. Expert system and fuzzy logic have their own significant capabilities the combination of both technologies that forms a fuzzy-expert system or a hybrid system could increase the systems performance (Herrmann, 1996). Due to the capability of fuzzy logic and expert system, pest activity prognosis in rice field using fuzzy expert approach developed to provide information to the farmers and researchers through the internet. Since rice is the main staple food of the Malaysian and Kedah is known as 'rice bowl' Malaysia, therefore this study focuses on the pest's activity in the rice fields. In MyPEST, the type of pest that causes damage to the rice plant is determined by the expert system. On the other hand, Fuzzy Logic approach is used to forecast the pest activity level. This is important so that early treatment or action can be applied before damage to the plant becomes worst. The system helps the user by managing the consultation which is performed by the expert system and fuzzy logic to make prediction and dealing with the natural and uncertainly data using linguistic variables. This web based application system also helps the farmers as well as agriculture institution representatives to manage farm successfully and to using more than one attributes involved, the less rigid 3 dimensional decision graph was produced. The identification for the type of pest is also involved in the first phase of this system which followed by the activity forecasting based on the identified pest. The system has been verified by MARDI entomologist and the system is confirmed to benefit the researchers at MARDI, MADA and DOA particularly, and the farmers at large.