Will the trace clustering algorithms be implemented on ProM6 (other than AHC) as in ProM5 ?

Martin_Prodel

Hello everybody,

Lately, I have been intensively reading articles about Process Mining and I went through Prom 5.2 and Prom 6. I am currently working in the healthcare domain, thus as described in Trace Clustering in Process Mining (2008, M. Song et al.), the use of trace clustering is of a major interest to reduce the high variability of the underlying process. 

To try the trace clustering algorithms (those described in the article) on my own data set, I first used Prom 5.2 and the "Trace clustering" plug-in. It allows us to choose among 7 different algorithms (all developed by Minseok Song):

- (New) Agglomerative Hierarchical Clustering
- Quality threshold clustering
- K-mean clustering
- Farthest first clustering
- Density based clustering
- EM clustering
- SOM

As I now went on ProM6, I saw that the GuideTreeMiner and the ActiTrac plug-ins are both providing trace clustering technics. Unfortunately, they only propose the AHC algorithms.

My question is : Is there any plug-in on Prom6 with the other algorithms, or is it planned to be added to the GuideTreeMiner?

How difficult would it be to implement the ‘Trace Clustering’ plug-in from Prom5 to Prom6?

 I have not seen any discussion about this matter, I hope my topic it is not redundant with any other.