Have you ever had to scroll through 100s of reviews on an e-commerce website and found yourself more confused than ever? Do you find the 5 star system arbitrary and too subjective? Did you wish there was some sort of summary you could read which tells you about, how people felt about the product? If yes, then you get the inspiration behind our app.
What it does
Riley is a chrome extension that takes all reviews about a product using Macy's review services and then use's IBM's tone-analyzer to analyze the review text. The tone-analyzer classifies the review sentences into five basic emotions anger, joy, sadness, disgust and fear. We display the number of review sentences that was found to be joyful, angry etc.
How we built it
The application uses Macy's review services API to get a list of reviews about a product. We then use IBM's tone-analyzer to analyze the emotions detected in the reviews. A NodeJS server makes all the API calls and acts as a RESTful server for our application. AngularJS along with chartJS was used for the frontend.
Challenges we ran into
We learnt a lot about CORS (cross origin resource sharing) today. Most, if not all of our problems centered around trying to get the chrome extension to accept the data from the API services.
Accomplishments that we're proud of
Getting a good, clean looking and fully functional app ready
What we learned
NodeJS, CORS, Chrome Extensions.
What's next for Riley
Currently Riley only works with the Macy's website, we would like to get this across to all e-commerce websites.