Vol. 7, Issue 1, Part K (2025)

Agricultural extension services are essential for enhancing farming techniques and productivity. Nevertheless, obstacles such as a shortage of available specialists, fluctuating guidance, and accessibility problems hinder the delivery of effective advisory services. AI-powered expert systems present a hopeful alternative by delivering immediate, data-informed suggestions for diagnosing and overseeing agricultural practices. This research investigates the usability, efficacy, and influence of an expert system designed to assist farmers through extension advisory services. A systematic interview framework gathered information from farmers with diverse backgrounds regarding system usability, accuracy of recommendations, user motivation, risk propensity, and technical abilities. To evaluate the influence of essential factors on the adoption and effectiveness of expert systems, we employed Multiple Linear Regression (MLR). The MLR model facilitated the assessment of how variables like farmers' education level, agricultural experience, system usability, and risk perception affected their utilization of the specialist system. The findings reveal that the majority of farmers regarded the expert system as user-friendly, efficient in delivering timely advice, and very precise in identifying agricultural problems such as soil health oversight, crop choice, and pest management. The MLR assessment suggested that educational background and prior experience with technology notably influenced adoption rates. Farmers possessing limited technical expertise required additional training to maximize the advantages of the system. Although the feedback was generally positive, several apprehensions were expressed regarding the system's capability to adjust to local circumstances and the need for continuous updates to tackle immediate agricultural issues.
This research underscores the capabilities of AI-based expert systems to revolutionize agricultural extension services. By reducing dependence on human consultants and delivering reliable, scalable, and precise guidance, these systems can significantly enhance 
agricultural efficiency. Nonetheless, advancements are required in interface design, region-specific content, and training programs for farmers to ensure broader implementation. Subsequent studies should concentrate on improving machine learning models, integrating adaptive learning techniques, and expanding access for smallholder farmers to widen the effect.