At the start of this Hackathon, RBC came to hackers for a solution to a task, they wanted to leverage the speed of information spread of social media to monitor, analyze, and categorize issues from customer feedback. Hoping to provide a compelling and real-life impacting solution, we decided to take upon this challenge. Also, the prize was also a very inspiring factor.
What it does
To meet their demands, we decided to create a hack that would be able to pull data from twitter, one of the most utilized platforms to exclaim one's right to freedom of speech and analyze these tweets to determine relevant information about the services and applications of our business of choice.
How we built it
Using the Twitter API and the tweepy python library, we are able to pull tweets that contain specific keywords such as the names of businesses such as RBC, Rogers, etc. From these tweets, we retrieved key information such as the text content of each tweet and the time when the tweet was posted and stored them into a JSON file. Then, we routed that file into a python script that used google's Natural Language API to analyze the sentiment of the tweets. After analyzing all of the tweets, we are able to provide a value that represents the ratio of negative to positive tweets which should indicate if there is a current problem with the various services or applications that the business provides.
Challenges we ran into
While we were grinding away at our hack, we were afflicted by a mental hurdle as the food provided was mediocre which induced us to order fried chicken that was made with real chicken. (Just kidding, the food was just subpar, but the carrots were pretty alright.)
Accomplishments that we're proud of
We were able to implement most of the functionalities that we initially intended and we drank 8 Red Bulls as a team.
What we learned
Too many Red Bulls may be bad for you. Twitter API does not take as long to acquire as we originally thought.
What's next for cuAnalyze
We would like to implement more social media platform from which we could pull data and further improve our user interface and experience. Furthermore, we want to be able to analyze the tweets with a list of keywords that help the business identify which service or application is the subject of the public's complaints.