Dear Quazy, 1) You can use the software that best fit to your requirements. Maybe, it could be Orange, R, IBM Statistics, SAS, Phyton libraries, etc. Even, you could use statistical approaches also, such as the correspondence analysis; 2) a software is not a methodology...take a look of traditional approaches before make your choice (e.g., crispdm, semma, kdd process, etc.); 3)https://doi.org/10.1007/978-3-540-30116-5_58