Starting Strong: Building the Trending Feature in SoundSoar

Author: Tori Grasso

Date: 9/1/2024

Introduction

Hi, I’m Tori Grasso, a passionate freelancer and a master's student working on an exciting capstone project called SoundSoar. My goal is to create an application that leverages machine learning to predict which songs will become trending hits based on their features. This project combines my love for technology and music, aiming to provide valuable insights for content creators, influencers, and small businesses. In this first blog post, I'll walk you through the initial steps of developing SoundSoar's trending feature, share the tools and resources I've used, and reflect on the challenges and successes of this past month.

Feature Development

SoundSoar is an innovative application designed to help content creators, influencers, and small businesses stay ahead of music trends. The app's core features include analyzing song data to predict which tracks are likely to trend based on various musical characteristics. By leveraging the Spotify API and advanced machine learning algorithms, SoundSoar provides insights into which songs have the potential to go viral. These predictions can help users enhance their content strategy by choosing the right music to engage their audience. Additionally, SoundSoar aims to offer a user-friendly interface and customizable dashboards for tracking trends and insights tailored to individual needs. See the high-level wireframe below:

This month, I focused on laying the groundwork for SoundSoar by setting up the web application on AWS Lightsail. I configured the basic navigation and user functionalities to create a seamless user experience. With these foundational elements in place, I have also made significant progress on the trending feature, which is about 60% complete.

The features that I chose for my prediction algorithm is as follows:

The next steps in the development process will involve starting the data tasks necessary to train the machine learning models that will drive the trending feature. This includes collecting and processing data from the Spotify API, which will be used to train the models. Once the data is prepared, I will move on to testing and fine-tuning the models to ensure they accurately predict trending songs based on various musical features.

Retrospective

What Went Right This Month?

This month, I successfully deployed the SoundSoar web application on AWS Lightsail and set up the foundational navigation and user interface. I made significant progress on the trending feature, bringing it to about 60% completion. The decision to focus on predicting streams rather than sentiment analysis has proven to be beneficial. It has allowed me to create a solid foundational piece for the application, setting the stage for further development and ensuring that I stay on track to complete the project on time.

What Went Wrong This Month?

The initial delay was due to my original plan to include sentiment analysis of tracks. After discussions with my advisor, we agreed it would be more feasible to scale back the project to focus on stream predictions. While this adjustment was necessary to meet the project timeline, it did cause some delays as I adapted to the revised scope. Technical challenges with AWS Lightsail also required additional time for troubleshooting and configuration.

How Can I Improve Moving Forward?

Moving forward, I will ensure to set realistic timelines and be prepared for adjustments in project scope if needed. Focusing on the core features first and ensuring they are well-developed will provide a strong foundation for future enhancements. I plan to start the data collection and preparation tasks earlier to avoid delays in model training and fine-tuning. This approach will help me stay on track and potentially allow for the integration of additional features, such as sentiment analysis, after the capstone project is completed.