[Pycon] [new paper] "Akilesh Lakshminarayanan" - Build your own chatbot using the Facebook Messenger API and Python-NLP packages

[info a pycon.it]

Title: Build your own chatbot using the Facebook Messenger API and Python-NLP packages
Duration: 240 (includes Q&A)
Q&A Session: 0
Language: en
Type: Talk

Abstract: While designing and developing your own functional chatbot might seem like a herculean task, python makes it extremely easy to add intelligent conversational functionality, and with pretty good accuracy! In addition to this, once you are familiar with Facebook messenger API (which also has a wide range of wonderful conversational interface elements) it’s actually not that hard a task to get your chatbot into production on Messenger.

The purpose of this tutorial is to guide one through all the whole process of designing and developing a conversational chatbot. The tutorial is broadly divided into 3 segments - 

Segment 1 (1 hour) - Introduction to Messenger API; using various conversational interface elements provided by the API (like sending images, links and quick replies); and setup simple chatbot application on _Heroku_, integrate with messenger.

Segment 2 (2 hour) - Adding NLP capabilities to the chatbot using _NLTK_ (for text classification) and _spaCy_ (for entity recognition, part of speech tagging) to enable appropriate response generation for various types of user queries.

Segment 3 (1 hour) - Introduce and setup webhooks; setup a Facebook product page, integrate messenger with your application and get your end to end bot into production!

Throughout the tutorial, I will be using Jupyter notebooks to introduce and explain various code segments. We will then write modular application code for each segment of the tutorial. This will be a code intensive tutorial, where participants would be encouraged to get their hands dirty writing their own code, through the entire process of developing a functional chatbot.

The takeaway from the tutorial would be for the participants to have an end-to-end working chatbot, that can logically answer user queries using NLP and some conversational interface elements (like quick replies) provided by the Messenger API. In addition to this, all the participants will have a complete picture of all the necessary tools to design and develop a chatbot, so that they can leverage it for their products.

Tags: [u'nltk', u'datascience', u'nlp', u'pydata']