Design and Implementation of a Semantic IoT-based Intelligent Q&A System for Greenhouse Agriculture Management

Greenhouse agriculture is an important contributor to food security and sustainable production. However, it is still difficult to manage environmental variables and make knowledge-based decisions for greenhouses. In the study, described the design and implementation of an intelligent Question & Answer document using Semantic Internet of Things (IoT). In the design of the Q&A document we focused on ontologies to semantically model greenhouse knowledge. From there we encapsulated a Natural Language Processing (NLP) module in our Q&A document that allow us to accurately parse and understand user questions. After user questions are understood, the document has built in semantic reasoning and SPARQL data querying that allows it to provide correct user answers based on context. In trial tests on test users, we showed that overall answering correctness increase, answering latency decrease, and that participants were satisfied and would use the Q&A document in their agricultural responsibilities. The Q&A document has achieved its intended purpose of providing better agricultural decision-space knowledge and a tool to assist decision-making.