DESQ is a general-purpose system for frequent sequence mining. It features a simple and intuitive pattern expression language to express various pattern mining tasks and provides efficient algorithms for mining.
K. Beedkar, R. Gemulla, W. Martens
A Unified Framework for Frequent Sequence Mining with Subsequence Constraints [pdf (author version)]
DESQ is open source and available on GitHub.
The source code for the version of DESQ used in the experimental study of our ICDM16 paper is available here.
Please contact us to obtain a copy of the datasets used in our experimental studies. For some of the datasets (e.g., New York Times), you'll need to have the appropriate licences.