cam-perry/summaraize

School project: YouTube comment analytics using IBM Watson API for sentiment & emotion categorization, jQuery and a Flask server

summaraize's Issues

Overall sentiment graphic display

Display the overall sentiment of the video as detailed in wireframe and/or hi-fi mockup

Design unification

• Header spacing on analyze page
• Make input group for comment summary coloured based on sentiment
• Make toggle a part of the input group for comment summary

Wireframes for comment summary

Collect Watson analytics on all comments

Run all comments through Watson Discovery, Sentiment Analyzer, Tone Analyzer to get all data required for analysis.

Able to pull in all of a video's comments based on a URL or ID

Use YouTube Data API to pull in all a video's comments and just log them to console to prove it works.

Can't press enter on main channel form

Hosting on bluemix

GetCommentsForVideo() seems to be crashing for videos with large #s of comments

To be fair I did try it on a vid with 6000+ comments. It worked fine for another video with ~10ish comments. @cam-perry

Flask server boilerplate set up

Use the Flask starter code and commit it to Github so we can start building backend functions

Display top entities in a list with their % positive and % negative

Analyze page blank when moving over

• Loading spinner on page load
• Insert titles and results in same function

summarAIze button doesn't scroll up to top of page on click

Calculate overall sentiment

Average all comments to get an overall sentiment and return this to the frontend.

Wireframe mockups

Calculations for entity summary

Using data created by Watson, identify top entities and score them on sentiment, prepare data format for passing back to frontend.

Input field and submit button for a video URL

High fidelity mockups

Customize Website (Beautify)

Calculations for comment summary

Using Watson analytics on comments, prepare comment summaries for front-end to display.