The Project

There are many resources out there to recommend music based on user's taste, taking into account their listening history, saved albums, mood, and activities. These algorithms are helpful, but sometimes, you want something completely different. Instead of using listening history and saved albums to recommend a song to listen to, What Should I Listen to Right Now recommends an album and a genre based on seemingly random interests and qualities that the user indicates throughout a series of questions, such as how they take their coffee and if they possess the cilantro-is-gross gene or not.

My role

I executed every aspect of this project, including background research, illustration, coding, and writing.

Step 1: The Questions

I wrote out a list of dozens of potential questions to use, whittled it down to seven, and wrote six answers for each, one to represent each of the six possible genres that the user could be recommended (country, hip-hop, rock, pop, indie, and dance). There was no obvious correlation between answer and matching genre, to ensure that the end result is not based on the user's usual taste.

Questions ask about personal taste as well as in-the-moment variables such as mood and phone battery, ensuring that the result is unique, and the user can repeat the process another time and get a different answer.

Step 2: Illustration

While the page layout varies from screen to screen, each page is similar in color scheme, design, and illustration type. I drew every screen in Adobe Illustrator.

Step 3: Code

Next, I imported everything into Processing and wrote the code to make it into an interactive module. On the final screen, the user is recommended a genre, and the module pulls data from the Last.fm API to display an album in the current top 200 most popular albums in that genre on Last.fm.