Sentiment Analysis of Farmer Reviews on Pesticides for Disease Management

In agriculture, managing plant diseases is important for keeping crops healthy and improving yields. Farmers use pesticides to control diseases, but choosing the right pesticide is often difficult because it depends on the type of disease, the environment, and the specific needs of the crops. This makes it hard for farmers to know which pesticide will work best. While there have been some methods to assess pesticide effectiveness, no projects focus directly on using farmer reviews to help with pesticide selection. This is a gap that needs to be filled to help farmers make better decisions. This research proposes a new solution by analyzing farmer reviews on pesticides using sentiment analysis. We use a deep learning model called Bidirectional LSTM (BiLSTM) combined with GloVe word embeddings that have been pre-trained. to understand the sentiment in the reviews. The BiLSTM model helps improve the accuracy of sentiment classification by considering both forward and backward text dependencies. Our model processes a dataset of disease names, pesticide names, and reviews, classifying them as positive, neutral, or negative. The results allow us to rank pesticides based on how effective they are for specific diseases. This approach gives farmers a better tool to choose the right pesticide for disease management.