“Pattern Mining :Theory and Practice” (textbook in Thai, with SPMF)

Hi all, this is to announce that a new textbook in Thai has been published about pattern mining, which includes many examples using the SPMF software. The textbook named “Pattern Mining: Theory and Practice” is written by teacher Panida Songram from Mahasarakham University (Thailand) and can be used for teaching or self-learning, for students or practitionners. I have known the auhor for many years and I am very happy that she let me host a copy of the book that you can download from this link:
Pattern Mining: Theory and Pratice (PDF, 14.2 MB),

The book gives a good coverage of pattern mining. It explains algorithms but also contains many practical examples about how to use SPMF. Some key topics in the book are itemset mining, sequential pattern mining and multi-dimensional sequential pattern mining.

That is all I wanted to share for today. If you can read Thai, I highly recommend to download this book. πŸ˜‰

Philippe Fournier-Viger is a computer science professor and founder of the SPMF open-source data mining library, which offers more than 170 algorithms for analyzing data, implemented in Java.

(Visited 115 times, 1 visits today)


“Pattern Mining :Theory and Practice” (textbook in Thai, with SPMF) — 4 Comments

  1. Dear professor Philippe,

    thank you for sharing this book. I seriously want to read this book. would you share this book in English language.

    habib un nabi
    student of computer science.

    • Good evening,
      Yes, I also think that the book looks very interesting but I also cannot read Thai. I think it would be awesome to have an English version. But I don’t know if the author has any plans to translate the book to English. I will let her know about the suggestion.

      Best regards,

  2. Thank you for your information.

    I have a question. Can we use HUPM in recommender systems the same way we use collaborative filtering ? Can HUPM be used in prediction in general?

    • Hi,

      Yes, patterns can be used in a recommender system. For example, if someone buys {a} and you have a pattern that tells you that many people buy {a,b} then you could recommend {b} to that person. However, this simple approach would not really consider the profile of the customers. It would be possible to use different kinds of patterns to do prediction.

      Related tot his topic, in association rule mining, several people have used the association rules to do classification.

      Best regards,


Leave a Reply

Your email address will not be published. Required fields are marked *