mlh-fellowship / 0.1.1-blm-tracker Goto Github PK

what additional features should we add to the frontend/ heatmap

We need to take the streamed tweets for the Twitter Streaming API and classify them as movement-related tweets, then use the relevant tweet locations to create a heat map.

• retrieve tweets in proper format from Twitter
• find pre-trained sentiment analysis model
• retrain model with tweets from initial tweet stream
• run collected tweets through model and classify tweets as relevant or not based on sentiment analysis

Update readme and have it reflect the hackathon submission document

• Name our project
• Figure out Twitter streaming API
• Implement database and caching system
• Scrub/ filter incoming data
  Concurrently:
• Design front end and integrate with streaming - Flask
• Implement Keras classifier for sentiment analysis
• Update heatmap and frontend

Create a front end that will be using Flask to create a heat map of relevant tweets.

• Get data from MongoDB instance via thread
• Make Flask implementation to stream its data to a Google Maps display

Currently working on implementing Tweepy to stream tweets to a python script, will update with progress

• Obtain API Keys
• Create external file with API keys and other static information (and add this file to the .gitignore)
• Get tweet stream functioning with test keyword
• Add real keywords to additional file
• Start working on pre-classifier data scrubber

integrate local mongoDB database to store Twitter streaming data.

Tasks:

• set up mongoDB locally
• connect with streaming API
• connect with front end heat map