Today, I am happy to announce that a new version of the SPMF data mining library has been released.
It introduces five new algorithms:
- the Carpenter algorithm (Pan et al., ICDM 2003), which is specialized for mining closed itemsets in transactions where the transactions contains a large number of items, and the number of transactions is relatively small. This is especially useful for biological data. The implementation is very efficient with multiple optimizations such as transaction merging.
- the Carpenter Max algorithm, which is a version of Carpenter for mining maximal itemsets by post-processing after applying Carpenter.
- The HMG_GA algorithms for discovering compressing sequential patterns in sequences using a genetic algorithm (M. Z. Nawaz et al., 2025)
- The HMG_SA algorithms for discovering compressing sequential patterns in sequences using simulated annealing
- And the GRIMP algorithm for discovering compressing itemsets in a transaction database using a genetic algorithm (M. Z. Nawaz et al., 2025).
Besides, some various other improvements have been made to the SPMF software, including a new tool in the GUI of SPMF called the Algorithm_Graph_Visualizer to visualize the similar algorithms in terms of input and output or categories as a graph (I had discussed about that in a recent blog post):
Besides, I have improved various graphical interface components such as the MemoryViewer, HistogramViewer, GraphViewer and SPMF Text Editor by fixing a few bugs and adding new features. I have also added a new mode in the TransactionDatabaseViewer to view transactions either as lists or as columns.
If everything goes, I might release more algorithms before the end of this year or early next year. I have several algorithms waiting to be integrated in SPMF!
To all users, thanks again for using the software! And to all contributors: thanks for your help!
If you have any algorithm that you would like to integrate in SPMF, feel free to contact with me by e-mail.
