Inspiration
Bruno studies, he has found a passion in making beats and original music. This lead to the inspiration to incorporate sound into this project.
Our team member Bruno's favorite hobby is to produce digital music. He thought it would be cool to turn data into sound.
What it does
The code is designed to allow you to obtain a list of tweets for any user and use this program to conduct a sentiment analysis. This analysis determines if the tweet was positive, negative, or neutral. We then start with a low base octave sound pitch and as the tweets polarity increases, decreases, or remains the same, so does the pitch of the sound. Ideally creating a musical expression of the tweeter's emotions.
How we built it
First of all the libraries that were used are: Gson, Sentiment140, JFugue and Apache HttpClient. The first step was to compile the tweets from a user from a CSV file. We turned this data into a JSON object, sent it to the API, and received back another JSON object with a specific data entry named polarity. This polarity is mapped to a 0, 2, or 4, which corresponds to negative, neutral, or positive SENTIMENT from the tweet as analyzed by the API. This data is then used to play a specific tone through JFugue depending on the emotion. Essentially, the next note that plays will be a lower or higher note than the previous one based on the emotion. The final output is a track of notes and a CSV file to analyze it with graphs.
Challenges we ran into
Group member bailed were left with two. Problems with JSON transformation. Creating a CSV file from data.
Accomplishments that we're proud of
Completing it, and successfully getting data from a server and analyzing it.
What we learned
How to communicate with a server and utilize that data
What's next for Tweet Sounification
Creating a better sounding track at the end, visualize the notes in a graph
Log in or sign up for Devpost to join the conversation.