It can be a good topic. But it is better if you try to add something new to this topic such as incremental skyline itemset mining
or skyline itemset mining in a stream or something else that has not been done before. Or otherwise to propose some new techniques or idea on skyline itemset mining.
You can get the source code of two skyline itemset mining algorithms
in the SPMF data mining library
, which is faster.
Both of these algorithm are for mining skyline itemsets when considering utility information.
You could do something related to that. Or you could do something different such as skyline uncertain itemset mining etc. But the point is that for a PHD you should make something new that is not trivial. So you have to think about the different possibilities to find something good related to that.
Edited 1 time(s). Last edit at 02/01/2017 04:39AM by webmasterphilfv.