Today, I am pleased to announce that a new version of the SPMF data mining software is released (v.2.67) and this is a major update. In this new version, I have improved several aspects of the user interface, and also 14 new pattern mining algorithms are added (thanks to Nabil Azizi, Chen Hui, especially). Besides, some previous algorithms have been improved with new features (in particular, improvements to cgspan were made by the original author Zevin Shaul).
The new algorithms are:
- the TriBackClo algorithm (Azizi et al., 2026) for frequent closed sequential pattern mining (thanks to Nabil Azizi et al. for providing the source code).
- the SNA-SPM (Ledmi et al., 2026), SUI (Huynh et al, 2022), and TreeMiner (Rizvee et al., 2020) algorithms for frequent sequential pattern mining
- the VEPRECO (Mordvanyuk et al., 2022) and DISC-ALL algorithms (Chiu et al., 2004) for frequent sequential pattern mining (thanks to Chen Hui for providing implementations).
- the FSG (Kuramochi and Karypis, 2004), FSP (Han et al., 2007), FFSM (Huan et al., 2003), and FSMA (Wu & Chen, 2008) algorithms for frequent subgraph mining (thanks to Chen Hui for providing implementations).
- the CloseGraph (Yan & Han, 2003) and CFGM algorithm (Peng & Zhang, 2023) for frequent closed subgraph mining.
- the MISAMiner_closed and MISAMiner_max (Wang & Yang, 2026) algorithms for frequent closed itemset mining and frequent maximal itemset mining, respectively.
Besides, SPMF 2.67, introduce a new launcher window like this with four options:
The top left option opens the traditional window of SPMF for launching a single algorithm.
The top right option is a novel step-by-step analysis interface where rather than selecting an algorithm first, the user selects a data file, and then can choose preprocessing algorithms and visualization before applying an algorithm. It looks like this:
The third option at the bottom left is the Workflow editor, which allows to chain multiple algorithms to create a workflow using a visual interface.
The fourth option is simply a drop-down menu offering access to other tools from SPMF.
Besides, there are new tools for visualization of patterns in SPMF such as the association rule explorer:
And there is a new Sequential Pattern Explorer tool for browsing sequential patterns using an intuitive interface:
This is just a brief overview of the improved user interface of this new version of SPMF. There are also many other smaller improvements.
Besides, as explained in a previous post, the input and output files of SPMF now have explicit file types, which allows a better handling of files by SPMF.
That is all for today. I have uploaded the new version, and now I will also upload it to Github and also update the different pages on the website as well over the next days.
Thanks again to all users and contributors of SPMF. If SPMF is useful in your research please support it by citing the research papers of SPMF:
- Fournier-Viger, P., Lin, C.W., Gomariz, A., Gueniche, T., Soltani, A., Deng, Z., Lam, H. T. (2016). The SPMF Open-Source Data Mining Library Version 2. Proc. 19th European Conference on Principles of Data Mining and Knowledge Discovery (PKDD 2016) Part III, Springer LNCS 9853, pp. 36-40.
- Fournier-Viger, P., Gomariz, A., Gueniche, T., Soltani, A., Wu., C., Tseng, V. S. (2014). SPMF: a Java Open-Source Pattern Mining Library. Journal of Machine Learning Research (JMLR), 15: 3389-3393.
