The premise behind our app was that people do want other people's opinions on places but don't necessarily want to read a bunch of reviews to get them. Some people understand things better visually, hence using Yelp's Reviews dataset we put together an app that passes those reviews to a sentiment analysis algorithm and pulls the keywords and sentiments of the reviews. With that data, we passed it to a data visualisations library called D3js which allowed us to create a bubble graph of the reviews. The size of the bubble represents the relevance of the keyword and the colour represents the sentiment.

Presentation :

Github :

"A picture is worth a thousand words"

  • Unknown
Share this project: