Song Recommendation Engine

A song recommendation engine built using Spark and Hadoop.

The model is optimised using the Alternating Least Sqaures (ALS) algorithm.

Dependencies

• Hadoop - 3.2.x
• Spark - 3.0.x

Usage

Extract the dataset

unzip dataset.zip

Copy the files to Hadoop

hdfs dfs -copyFromLocal 10000.txt /user/<name>
hdfs dfs -copyFromLocal song_data.csv /user/<name>

Run the Spark job

$SPARK_HOME/bin/spark-submit rec_engine.py

Contributors:

• Roshan Daivajna - PES2201800372
• Nikhil KR - PES2201800044
• Ruchira R Vadiraj - PES2201800602