Domain-Specific vs. General Purpose Large Language Models in Agriculture: A Narrative Review of Retrieval Augmented Generation (RAG) for Evidence-Based Extension Services
Abstract
The rapid development of Large Language Models (LLMs) has enabled the use of chat-based systems to support agricultural knowledge sharing, particularly for rural communities. However, general-purpose models often lack domain grounding, which can lead to inaccurate outputs and a localization gap where recommendations do not match local conditions and available resources. This study presents a narrative review guided by PRISMA methodology, analyzing literature published between 2018 and 2026 from platforms such as Google Scholar, ScienceDirect, IEEE Xplore, and MDPI. The review focuses on how LLMs are applied in agricultural extension services and examines their limitations in real-world use. The study compares general-purpose models with domain-specific systems such as AgriGPT, AgroLLM, and AgriBERT. The findings indicate that domain-specific models perform better in terms of accuracy and relevance when trained on curated agricultural data, while general-purpose models are limited by lack of contextual grounding. Retrieval-Augmented Generation (RAG) is identified as an effective approach for improving reliability by linking outputs to verifiable sources. In addition, multimodal systems enhance performance in tasks such as crop disease detection. The reviewed literature suggests that integrating RAG-based systems may improve traceability, grounding and reliability in agriculture, particularly where evidence based extension support is critical.
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.