Scholarlink

India’s agricultural productivity is hindered by climate variability, fragmented landholdings, and limited access to localized, data-driven advice. Traditional crop recommendation systems often fail to address diverse agro-climatic needs. We introduce Agri Buddy, a multilingual chatbot that uses large language models (BERT and T5) to provide personalized crop suggestions via natural language. To fine-tune the models, synthetic dialogues are generated from structured agro-climatic data. Achieving 98% accuracy, Agri Buddy outperforms existing methods and demonstrates the potential of LLM-based dialogue systems for accessible, intelligent agricultural support.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Agri Buddy: From Queries to Crops with a Context-Aware Agricultural Chatbot

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Somrita Sarkar,
  • Prasenjit Betal,
  • Pabitra Mitra,
  • Anupam Das,
  • Mrinal Jha,
  • Prashant Bisht,
  • Sarthak Sablania,
  • Thakare Vedant Sharadrao

摘要

India’s agricultural productivity is hindered by climate variability, fragmented landholdings, and limited access to localized, data-driven advice. Traditional crop recommendation systems often fail to address diverse agro-climatic needs. We introduce Agri Buddy, a multilingual chatbot that uses large language models (BERT and T5) to provide personalized crop suggestions via natural language. To fine-tune the models, synthetic dialogues are generated from structured agro-climatic data. Achieving 98% accuracy, Agri Buddy outperforms existing methods and demonstrates the potential of LLM-based dialogue systems for accessible, intelligent agricultural support.