Determining What Elements of a Song Make it Popular Based On Trend Studies and Clustering Methods by Logan Wrinkle and Mahim Mathur
Our dataset contains 2000 of the most popular songs on Spotify over 20 years (100 for each year), which tracks not only genre but elements like energy, danceability, upbeatness and acousticness for each song.
Our group wants to explore two questions:
- What trends can we spot over the 20 years that the dataset covers?
- What elements, or combination of elements, is associated with popular songs?
In this repository, we have our Jupyter Notebook that includes the code we used to analyze our dataset and generate visualizations. We also included our poster that summarizes our findings. In the 'graphs' directory, we have some of the images generated by our code.