Email: daranrat.j@cmu.ac.th

Affiliation: Faculty of Agriculture, Chiang Mai University, Thailand

Nationality: Thai

Analysis of User Engagement and Sentiment in an AI-Powered LINE Chatbot for Agricultural Learning in Northern Thailand

Abstract

This study examines the implementation of an AI-powered chatbot developed on the LINE platform and integrated with OpenAI's GPT model to support agricultural learning among university students and farmers in Northern Thailand. Designed as part of a participatory and action-oriented initiative, the chatbot fosters flexible, dialogic, and context-aware learning through natural Thai language interaction. Rather than relying solely on traditional survey-based evaluations, this study employs sentiment classification to assess learner engagement and perceived satisfaction. A combination of keyword heuristics and supervised machine learning techniques was used to categorize learner-generated messages into sentiment labels (positive, neutral, and negative). This method enables a behavioral understanding of how students and farmers interact with the chatbot in real time, without the need for structured questionnaires. The system was piloted in both formal undergraduate agricultural courses and informal community-based settings, including early-career and practicing farmers. Conversation data were collected and anonymized for sentiment classification and interaction pattern analysis. While initial findings suggest that clear, contextualized responses are often associated with positive sentiment, the study also acknowledges limitations in user diversity and message volume. As the pilot was implemented with a relatively small and specific learning group, further research is needed to validate these insights across broader agrarian populations and educational contexts. The use of sentiment-based behavioral data offers an alternative to top-down evaluation methods and reflects how learners in agriculture engage critically with emerging technologies. The chatbot was designed as a dialogic and participatory learning tool, supporting more flexible engagement in agricultural education. While grounded in practical deployment, The project also invites reflection on broader questions of how digital technologies may reshape knowledge production, inclusion, and agency within rural learning environments. In this sense, the study resonates with concerns shared by critical perspectives in agrarian studies, particularly those interested in the evolving relationship between technology, education, and the conditions of agrarian life.

Bio

Dr. Daranrat Jaitiang is a full-time lecturer and early-career researcher at the Faculty of Agriculture, Chiang Mai University, Thailand. She earned her PhD. in Agribusiness Management in 2022 from the National Pingtung University of Science and Technology in Taiwan. Her research focuses on sustainable food systems, agricultural education, rural development, and digital inclusion in Northern Thailand. Her work lies at the intersection of agricultural development and digital transformation, with an emphasis on how participatory technologies can support farmer-centered learning. She currently leads a project that integrates LINE-based AI chatbots into agricultural coursework to support students and young farmers. The chatbot system, powered by generative AI (GPT-4o), is designed to enable multi-turn dialogue, contextualized learning, and inclusive access to knowledge in rural settings.

Dr. Jaitiang adopts an action research approach and engages with critical perspectives in agrarian studies, participatory pedagogy, and digital sovereignty. Her research includes innovation mapping using Q methodology, PM2.5 mitigation, leadership development in highland areas, and bioeconomy value chains. Her academic work reflects a balance between technical innovation and equity-oriented inquiry into agrarian change. As an early-career scholar-activist, she is committed to reimagining how agricultural knowledge is produced and shared. Her work centers rural learners, promotes inclusive, context-responsive education, and contributes to regional efforts toward agrarian futures in Southeast Asia.