Medical check-ups, illness diagnoses, and treatment suggestions are the most common outcomes of a visit to the hospital. The vast majority of individuals all across the globe have engaged in this activity. Many people believe it to be the most accurate method of determining one's health. Nowadays, individuals are less conscious of their health. In their hectic schedules, people often overlook the need to keep their health in check. The use of Natural Language Processing (NLP) methodologies and their use in constructing conversational systems for health diagnosis boosts patients' access to medical information. Medical chatbots have been developed and applied in various clinical settings to provide conversational tools accessible to a broad range of healthcare providers and patients. In this suggested system, a medical chatbot is designed to be a conversational agent that pushes patients to discuss their health difficulties based on the symptoms presented; the chatbot delivers the diagnosis. Preparation of text-based documents, document tagging, symptom detection, and illness prediction is part of the proposed Chatbot-Based Medical Diagnosis (CBMD) system. The CBMD method encourages people to open up about their health concerns, gives a proper diagnosis, and prescribes treatment. NLP is used in this instance to do text processing. The first step is for the patient to type in their symptoms. All the text in those files has already been preprocessed using techniques like stemming, stopping, and tokenization. The knowledge source's useful content is labeled to identify quality information from a single document. For illness diagnosis, the patient's symptoms are compared to those in the knowledge base. Finally, Deep Neural Network (DNN) classifier is used to classify the patient's symptoms matched to their corresponding ailment. DNN recognizes the signs and symptoms based on the patient's input. To help patients, DNN analyzes their symptoms and provides the most appropriate remedy. In the last step, the outcomes of the classifier are evaluated using several metrics, such as precision, recall, f-measure, and accuracy.