• Version
• Download 3
• File Size 380.63 KB
• File Count 1
• Create Date 30 January 2025
• Last Updated 31 January 2025

Developing a Machine Learning based Humanized Music Recommendation System

Mrs. P. Kamakshi Thai*1 , Chirag Tiwari*2 , Vishwanadhula Divya*3 , Mulamalla Thulasidhar Reddy*4 *1Assistant Professor, Department Of CSE (AI&ML) Of ACE Engineering College, India.

*2,3,4Students of Department Of CSE (AI&ML) Of ACE Engineering College, India.

 

Abstract: Music has become an integral part of our lives, with platforms like Spotify and Amazon Music dominating the market. However, they still often fall short in terms of user satisfaction. This research aims to focus on matching the energy of the song played by the user instead. This research aims to create an alternative model that does not focus on traditional features such as popularity or artist but on nuanced features such as energy, liveness and tempo to add the specific ‘human touch’. To do this, it uses a Spotify dataset from Kaggle which has a wide range of attributes such as acousticness, danceability and valence, among others. By using these underutilized features, this research aims to create a recommendation system that provides more personalized and satisfying music suggestions. This research will include of making a web application which will allow users to play music and this model and web application will be deployed using AWS.

Keywords: Personalized song matching, nuanced features, human-touch, user satisfaction enhancement, Data preprocessing, Clustering Analysis, Dynamic playlist generation

Download