PyCon Balkan 2019 Talks
TLK PyCon Talk
NLP for Dinner - Chatbot language learning
I used Python + Rasa NLU powered by Tensorflow and Recurrent neural Networks architecture to create a bot that understands and texts in Czech. Why ? The bot created with Python + RASA + NLP Techniques is currently helping me to learn a new
language and prepare myself for real case situations, such as order food in a foreign country. The domain of the chatbot so far is Order food so my Waiter bot has a real impact in my life, because being Colombian and living in Prague, the language is a
big challenge, I took Czech lessons for beginners and rapidly I found out that the traditional way of teaching, prepares you for Saying things, but does not train you to process the reply and talk back, so I used Tensorflow and Built a Chatbot that simulates being a waiter and 'speaks' Czech to me.
I use it from time to time as a Sparring partner and now I am able to order food all by myself, thanks to my Bot trained in Czech. With my bot who speaks Czech, I simulate scenarios in which I visit a restaurant and the bot process my order if I text back in Czech the proper entities.
The audience can benefit from this talk in a way that they can see through my use case how they do the same in their lives. Learn about NLP beyond the hype and see the concepts applied to something that can relate to many of us: Starting a new life abroad.