PrayerBot

Inspiration

We believe that prayer is powerful and effective. Many of us fail to pray as much as we say we will. Many of us forget to pray for those we've told we'd pray for them.

What it does

This classifier determines whether a string of text is talking about prayer or not.

How I built it

This was built with Python in Jupyter Notebook with the help of pandas, numpy, nltk, scikit-learn, and lots of help from stack overflow and the organizers of the hackathon.

Challenges I ran into

Webscraping various pages with different layouts, tokenizing the string, determining the right classifier to use and how to make it better.

Accomplishments that I'm proud of

I'm proud to have learned webscraping and more about natural language processing!

What I learned

I learned a lot about natural language processing and what it takes to see a project come to fruition.

What's next for PrayerBot

• getting more data (texts) to use to train the classifier with the help of APIs.
• integrate into Echo

Built With

• nltk
• numpy
• pandas
• python
• scikit-learn