- Mysql server 5.0 or later
- Java version 6.0 or higher
- Libraries required
- mysqljavaconnector5.1.26.jar
- jsonsimple1.1.jar
- jdf5.jar, jdf5obj.jar to read hdf5 files
-
You need to download and Install mysql MySQL > http://www.mysql.org
-
Configure database properly and make sure its secured.
-
The database server uses root as default username and password. Password can be changed by using UPDATE mysql.user SET Password=PASSWORD('cleartext password') WHERE User='username' AND Host='%.host'; Also a new user can be created and should be granted insert, select and update permissions You need to change the username and password in DBConnect.java according to the DB user and password.
-
Create database schema musicanalytics and tables by running the MusicAnalyticsDB_Script.
-
Once schema and tables are created the data can be populated by using the java programs contained in this source code folder. This step have been completed.
-
The Source files consists of reading the raw data in various formats such as json, hdf5 and converts it to csv files using java. The csv files are then read, parsed and the data is cleaned and inserted into mysql using java jdbc connection. This step is also completed.
-
Steps 5 and 6 can be avoided by directly importing the data sql file provided into the database.
-
Database will then be populated and will be ready for data mining.
-
To run the user Interface, class MusicUI inside the analysis package needs to be run. This will display a user interface with list of users with prev, next and generate buttons. Click on prev or next buttons to get more users.
-
Select a particular user and click generate recommendation. This will display the listening history of user and the recommended songs suggested by the algorithm.
-
Select any song and click play button, this will update that song in user listening history in DB. If the user is listening that song more than once, the playcount will be incremented.
-
Play any other songs or select any other user and generate recommendation.