1
$\begingroup$

The FP-Growth Mining Algorithm finds frequent itemsets i.e. it does not search for patterns by item. If I want to find ALL frequent itemsets for an item, I can use the nodelink property to traverse all paths in the fp-tree for that item, which means I have to traverse all paths for all children of that item.

In the worst case this effectively means I have to run the mining algorithm on the entire fp-tree (and then filter the results for the specific item I'm looking for). This seems like an expensive operation to perform repeatedly, as I'm only looking for a small subset of results. Is there any other mechanism that would allow me to mine patterns selectively?

1) Mining Frequent Patterns without Candidate Generation; Jiawei Han, Jian Pei, and Yiwen Yin

$\endgroup$
1

1 Answer 1

0
$\begingroup$

It sounds like you need a second data structure like an index. An index tracks important metadata information about the data.

Graphs are typically not structured for fast querying. Indexes can be structured for fast querying.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.