My friends really like trying out different foods, so I wanted to give them something that would help them find their new favorite food place using Yelp reviews and the sentiments that they provoke, from joy to trust.

What it does

The Yelp sentiment analyzer takes in parameters such as the location you are at, the type of food you would like to try out, and the distance that you are willing to travel. Then, it outputs food places in that category along with the reviews and analyzes the overall sentiment and vibe so you don't have to. This would come in handy it you are in an unfamiliar area and are looking for trustworthy coffee shops to stop at. By getting the overall positive sentiment from the analyzer, you will be assured that you are in the clear.

How we built it

I built it in google colab using python, Yelp API, and the NRC Word-Emotion Association Lexicon (aka EmoLex). The EmoLex basically breaks up the sentence into words and then gives each word a sentiment. If the overall sentiment of the combined words (sentence) is a specific emotion, that emotion will be output.

Challenges we ran into

I realized that the Yelp-API only allows a developer to call 3 reviews with a limited amount of characters. This would mean that the Yelp sentiment analyzer might be biased and largely skewed as it only gets to analyze three reviews. However, I am looking forward to solving this by implementing a Twitter API that analyzes tweets about restaurants. This will help present a more holistic review of the restaurants.

Accomplishments that we're proud of

I am glad to have learned Postman because it helped me understand what parameters I needed to accomplish this project. I am also really surprised that I could figure out how to analyze the data using a the EmoLex.

What's next for Yelp Sentiment Reviewer

I will be trying to make this a user-friendly application as it is not the prettiest right now. I will also implement the Twitter API.

Share this project: