AN AI-POWERED CHATBOT FOR CONTEXT-AWARE FERTILIZER RECOMMENDATIONS USING HYBRID MACHINE LEARNING

In contemporary farming, proper and effective usage of fertilizers is very important in accomplishing maximum crop yields without destroying the environment.Conventional means of fertilizer recommendations usually involve the use generalized rules and guidelines, which neglect the field specific parameters therefore leading to an imbalance of nutrients, degradation of soil and poor yield [1].To overcome this problem, this study suggest an intelligent Fertilizer Recommendation System (FRS) that would implement machine learning to generate custom fertilizer recommendation depending on actual time agricultural inputs.The device is trained using a Random Forest Classifier based on an extensive training dataset of nitrogen (N), phosphorus (P), potassium (K), temperature, humidity, pH level, soil type, moisture, rainfall, crop type features.The model demonstrated predictive performance as its accuracy was 91.3%.The software product is designed on the basis of a modular architecture in the form of a Flask-based Python back-end, an HTML5-CSS3-JavaScript front-end, and a lightweight AI chatbot that is embedded to facilitate user interaction and query resolutions [6].The trend of historical predictions is displayed by the Chart.jsscript, and the input and output data of the user are encrypted in an SQLite database.The project will not only benefit the emerging industry in smart farming but also fill the digital divide in that the farmers will have a convenient and interactive, and intelligent agricultural decision support system.It is scalable and future-proof with scope of integrating it with IoT sensors, weather API and NLP-based multilingual chatbots [5][9].