Using NLTK & SciPy to build a small task bot accepting natural language

Objective

We will be building a small RESTful bot in flask which performs simple tasks on Natural Language queries like "Make a Note", using nltk,scikit-learn.

Whats for the listener :

• A real world use case and example of nltk and machine learning in action
• A high level understanding of how these systems work
• The code methodology can be expanded to make self versions of siri (small case)
• A rather not so booky approach on ml

Description

Almost everybody today wants to get their share on natural language processing, machine learning etc. We are on the stage where practical applications of systems using natural language as input and machine learning and their own intellect to find solutions upto some extent.

This talk will deal with some higher level logic involved in building a bot which is able to perform simple tasks like

• Taking a Note,
• Sending a message,
• Showing a contact etc.

The interface will be done in web but we can abstract that out since our apis will be RESTful :-)

Requirements

• Basics of Python.
• Some knowledge of Linux is appreciated.
• Pre-installed nltk, scikit-learn would be amazing :D
• Going through http://nltk.org/book/ it would be really appreciated but necessary.

Speaker bio

I love Automation.

I love Python.

I love Sharing and Learning.

I love the web-stack.

And

I love to do this all together!

I am currently a developer at MidNode and a founder of a startup which deals primarily with natural language and machine learning, Our product is in pre-alpha and we are planning for a mid-august / september launch.

Our whole techstack is python based (infact the same stack i would be discussing in this talk) and hence I will be giving some tips and suggestions on common pitfalls aswell.

In past I have worked on projects like Pixter and Colorbay at MetaRain in past.

Some Cool stuff i did with web-stack :
http://codepen.io/darkyen/

Links

• Coming Soon - Live Demo Link

Comments

• ▲
  1
  ▼
  

  [-][+] Aravind Krishnaswamy 284 days ago

  The outline looks interesting. What are some of the takeaways people attending can expect?

  [reply] [link]
  • ▲
    1
    ▼
    

    [-][+] Abhishek Hingnikar 283 days ago

    Adding.

    [reply] [link] [parent]
  • ▲
    1
    ▼
    

    [-][+] Abhishek Hingnikar 283 days ago

    Please have a look now :-)

    [reply] [link] [parent]

Login with Twitter or Google to leave a comment →