Towards Ontology-based Expert System Development and Evaluation for Rice Disease Identification and Control Recommendation

Various knowledge related to rice cultivation has been widely published on the web. Conventionally, this knowledge is manually studied by end users for use on rice diseases and pests identification in order to prevent the production losses. Despite its benefits, the knowledge has not yet been encoded in a machine-processable form. We improve this gap by turning the unstructured or semi-structured knowledge of rice diseases and their controls into the structured ones by employment of ontologies and semantic technologies. We externalize knowledge from existing reliable sources only. As a result, the developed ontologies offer axioms that describe abnormal appearances in each rice disease (and insect) and its corresponding controls. We also develop an expert system called RiceMan based on our ontologies to support technical and non-technical users for identification of diseases and insects from their observed abnormalities. We also introduce a composition service to aggregate users' observation data with others for the possible spreadable diseases and realize the controls. This composition mechanism, together with ontology reasoning, lies at the heart of our methodology. Finally, we evaluate our methodology practically with four types of stakeholders in Thailand, namely expert agronomists, non-expert agronomists, agricultural students, and ontology specialists. Both ontologies and RiceMan application were evaluated to ensure its usefulness and usability in various aspects. Our experimental results show that ontology reasoning is a promising approach for this domain problem.