Designing an Intelligent Chatbot with Deep Learning: Leveraging FNN Algorithm for Conversational Agents to Improve the Chatbot Performance

This research paper explores the design and implementation of a chatbot developed by students using deep learning algorithms. The primary focus is on employing a feedforward neural network, specifically a multi-layer perceptron, to create an intelligent conversational agent. Natural language processing techniques, such as lemmatization through SpaCy, are employed for efficient text processing. The dataset, extracted from a JSON file containing intents, patterns, and responses, is preprocessed using tokenization and transformed into a Bag of Words (BoW) representation. The training process involves epochs and batch size tuning, with model performance monitored through the TensorBoard and enhanced using callbacks, including ModelCheckpoint and EarlyStopping. The final model is saved and loaded for efficient utilization. The research extends to developing a user-friendly interface using Tkinter for seamless interaction with the chatbot.