hackathon-chat

recapbot

RecapBot

Problem

Have you ever been in a group chat and want to be able to filter through the relevant information? Our recapbot uses a machine learning algorithm (Bayes filter) to classify relevant from nonrelevant sentences over the Cisco spark API.

Tech Stack

Spark Streaming
Python for the backend *Naive Bayes filter

Setup

A bot was created in spark and an ngrok tunnel was used to the webhook. A learning algorithm classifies the sentences and the size of the text it altered (the larger the font, the more relevant the information).

Built With

cisco-spark
json
python

Submitted to

TechCrunch Disrupt London
- Winner Cisco Spark

Created by

I worked on python script that preprocessed chat messages and trained a Naive-Bayes classifier. I also wrote the python function that applies the classifier to the messages and modifies their markdown based on how relevant it is.

Aydan Gasimova
I implemented a python script that dealt with user input from Spark rooms and to process parameters passed alongside the calling of the bot.

Adrian C
Jason Plumb
Nicholas Keatley

Updates

Jason Plumb started this project — Dec 04, 2016 03:29 AM EST

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.