A study of conversational agent solution technologies for banana farmer assistance

Modern agricultural extension services help as a source of information for farmer queries. One of its applications are local hot lines where expert agents in specific crops or fields assist in farmer agricultural practices and decisions. As an already established technology in the field of customer services, conversational agents can help this solution by covering for experts when users have difficulties contacting them. These artificial agents can answer general queries at any hour, increasing the extension service accessibility. The creation of a contextualized solution is possible with the advancements of Natural Language Processing (NLP) and its accessibility in cloud services. We write this paper to collect state-of-the-art research on agricultural conversational agents and tools used for its design and deployment. General knowledge on current applications is gather and can later be used to help in the design of a localized solution when agricultural services are difficult to access. Lastly, we describe and analyze a chatbot prototype in the specific field of banana farming using IBM Watson services and the messaging platform Telegram App.