Music Recommendation System
Team : Aarthi Alagammai , Anupriya Saraswat , Divyanshu Malik, Garima Sharma , Harshit Rai , Nidhi Singh, Snehal Lokesh
Introduction: Recommender System are widely used today in all most all the applications.The purpose of a recommender system is to suggest users something based on their interest or usage history.Two most ubiquitous types of personalized recommendation systems are Content-Based and Collaborative Filtering. Collaborative filtering produces recommendations based on the knowledge of users� attitude to items, that is it uses the �wisdom of the crowd� to recommend items. In contrast, content-based recommendation systems focus on the attributes of the items and give you recommendations based on the similarity between them. We have created a Recommender sysem using Spotify We have Scrapped dataset from SPOTIFY using our custom scraper, "Scrapify". The Scrapped data is converted to as csv file and used for further processing.The dataset contains appromixately 11k observations
Data Description:
-name : Name of the user
-artist : Name of the artist
-danceability : Ranges from 0 to 1
-key : Ranges from 0 to 11
-mode : Ranges from 0 and 1
-instrumentalness : Ranges from 0 to 1
-duration : Duration of the song in minutes
-energy : Ranges from 0 to 1
-loudness : Float typically ranging from -60 to 0
-speechiness : Ranges from 0 to 1
-acousticness : Ranges from 0 to 1
-tempo : Float typically ranging from 0 to 150
-liveness : Ranges from 0 to 1
-valence : Ranges from 0 to 1
-popularity : Ranges from 0 to 100
-hollywood : Hollywood song 1 | Bollywood song 0
Project Goals
The goals for this project are:
-Scrap the website and collect the required data
-Organise the data into a Structured format
-
Gather insights from data analysis about the columns used
-
Perform EDA and remove unwanted columns
-Use the Cosine Similarity to calculate a numeric quantity that denotes the similarity between two songs. Since we have used the vectors, calculating the Dot Product will directly give us the Cosine Similarity Score.
- Ouput the top 5 recommended songs
Technologies Used:
- Python
- Google Colab
- Spotify API & custom scraper
Model Training Process
For model building we exclusively focused on using collabrative filtering approach because:
- We dont have any exclusive ratings of the users for the songs
- We just have user prefernces