Multi-level Closed High-Utility Pattern Miner

• src: Source code of the algorithm MLC-Miner and CHUI-Miner*

• taxonomy: synthesized taxonomy of the databases used in the work. Databases with real taxonomies can be obtained from the SPMF website.

• Java 8.
• An IDE (such as NetBeans or Eclipse) to import the whole project

• Within each algorithm, there is a test drive called TestXXXX.java.
• Changing database: replace the value of the dataset variable with your desire dataset. The driver will automatically append the _tax.txt and _trans.txt suffixes.
• Each databases must comes in two separated file: one containing the transactions, which name ends with _trans.txt. The other contains the related taxonomy of items, which name ends with _tax.txt. These two files should be place on top level of the package folder structure, outside the src folder. The format of both transaction database and the taxonomy strictly follow the SPMF format.
• Use the minutil variable to specify the minimum utility threshold. The values should be the exact threshold to be used, no relative (percentages).

The source code is released under GNU GPLv3.