Title: Sentimental analysis to enhancing agricultural productivity and sustainability

Authors: Puspalatha Chittem Setty; Bhadrappa Haralayya

Addresses: Visvesvaraya Technological University (VTU), Belagavi 590018, Karnataka, India ' Department of MBA, Lingaraj Appa Engineering College, Bider 585403, Karnataka, India

Abstract: In the era of mobile social networks, sentiment analysis has become a crucial tool for understanding public opinion across various domains, including agriculture. This study presents an advanced bidirectional encoder representations from transformers (BERT) model designed to analyse sentiment trends within the agricultural market, supported by machine learning (ML), deep learning (DL), and internet of things (IoT) technologies. By evaluating post-purchase reviews and textual data categorised as positive, negative, or neutral, the model captures valuable insights into consumer perceptions and emotional responses. These findings assist farmers, buyers, and producers in improving product quality and market strategies. Additionally, the study assesses agricultural productivity and performance metrics using the BERT framework, demonstrating its superiority over existing ML and DL models in sentiment classification accuracy and reliability.

Keywords: deep learning; convolutional neural networks; CNN; bidirectional encoder representations from transformers; BERT; recurrent neural networks; RNN; capsule networks; deep belief networks; DBN.

DOI: 10.1504/IJAITG.2026.153495

International Journal of Agriculture Innovation, Technology and Globalisation, 2026 Vol.5 No.2, pp.182 - 215

Received: 20 Jun 2025
Accepted: 25 Nov 2025
Published online: 11 May 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article