Chatbot_main.py is the final completed version of the code Developed using Python (chatbot code) and java (dataset) You'll need to install ALOT of libraries - tensorflow, pickle, json, numpy .... so be careful about that If you have issues installing it using the normal "pip install 'module' " command , then try using "python -m pip install --upgrade 'module' " 2 datasets have been uploaded - a) related to PESU academy b)related to Newolf Society, you can create your own dataset based on the type of service your chatbot is providing
#https://www.oracle.com/in/chatbots/what-is-a-chatbot/
#https://towardsdatascience.com/how-to-build-your-own-chatbot-using-deep-learning-bb41f970e281
#https://towardsdatascience.com/how-to-create-a-chatbot-with-python-deep-learning-in-less-than-an-hour-56a063bdfc44
#https://medium.com/analytics-vidhya/building-a-simple-chatbot-in-python-using-nltk-7c8c8215ac6e
You could even use java or html for the front end ,we've made it using python's gui framework Try using lemmatisation instead of stemming , try different types of stemmers and training models to figure out which one works the best If your chatbot's accuracy is around 0.9something it's good to go , if you get more than 1 or less than 0.8something then you'll need to try increasing the accuracy by adding more intents (this is the easiest way tho) It may seem hard in the beginning but it's not once you understand what's actually going on :) Also we could'nt upload the entire video of it working , so we uploaded just a small clip of it