Fresh data mining papers

Andrea and I just finished two papers that will appear at ICDM in Sydney, in December:

The first paper presents the first efficient algorithm for mining highly similar pairs directly out of a stream of transactions (where there is no way of storing all data, but we need to process data as it passes). The second proposes a new sampling method that can be used to extract frequent patterns from event sequences. It can also be seen as a graph mining result, that works on directed acyclic graphs. In the mean time, other miners are about to surface after a mining procedure that took way too long…

0 Responses to “Fresh data mining papers”


  1. No Comments

Leave a Reply

You must login to post a comment.