Skip to content
Conference Open access

Explainable AI-based Multilingual Crop Recommendation System for Precision Agriculture using Machine Learning

Oct 2026 · International Conference on Innovative Mechanisms for Industry Applications · pp. 48-56 · 0 citations · 10 references

Abstract

The problem of selecting crops using soil and environmental factors still poses challenges for precision agriculture. It directly impacts the performance of agricultural production, available resources, and the income of farmers. An intelligent multilingual crop recommendation system was developed using machine learning algorithms and Explainable Artificial Intelligence (XAI). This system will provide real-time decision support for the agricultural industry. We utilized the Kaggle Crop Recommendation Dataset that contains balanced samples of 2200 crops across 22 crop classes, and we used a total of 7 points of soil and climatic input data: Nitrogen (N), Phosphorus (P), Potassium (K), temperature, humidity, pH, and rainfall. We tested 8 different machine learning algorithms to develop a model for crop recommendation: Random Forest, Decision Tree, Naive Bayes, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Logistic Regression, XGBoost, LightGBM. Each model was evaluated by using stratified train/test splits and 10-fold cross-validation. The Random Forest classifier has a macro-averaged F1-score of 0.991 and achieves an accuracy of 99.09% on test data with a ± 0.38% cross-validation error; overall, results are very stable across folds. Feature importance analysis using Explainable AI techniques allows for improved interpretability and transparency in determining which environmental and soil parameters have the greatest influence on the suitability of specific crops. The Random Forest algorithm is deployed via a real-time interactive web application built on the Streamlit framework, with multilingual (English, Hindi, Telugu) capabilities to improve access for farmers who do not speak English fluently. The system features SHA-256 based user authentication, personalized history of crop recommendations, top 3 choosing crops with confidence scores, and crop-specific guidance for farming. The feature importance analysis indicates that rainfall and humidity are the most important factors predicting suitability for different crops. While the system showed excellent performance on benchmarks, it still requires validation against larger-scale, real-world agricultural datasets to demonstrate generalizability across various practical settings for commercial use.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.