search

uma-pi1 / mgfsm

Large scale frequent sequence mining
Apache License 2.0
42 stars 21 forks source link

Output with supporting sequence IDs #2

Closed kleii closed 5 years ago

kleii commented 5 years ago

I was browsing through the code but could not really pinpoint the place where I could potentially collect the supporting sequence IDs for each frequent pattern (and output them).

Is this possible to do in the algorithm (without too much effort)? If yes, where could I potentially start?

kbeedkar commented 5 years ago

It will be quite some effort if you are running MGFSM in distributed mode. In the sequential mode, its quite straightforward: In BfsMiner, you can get sequence ids from the posting list of each frequent sequence.

kleii commented 5 years ago

I was in fact interested in the distributed mode. Thanks anyway.