?
Leveraging Super-Scalarity and Parallelism to Provide Fast Declare Mining without Restrictions
UnconstrainedMiner is a tool for fast and accurate mining Declare constraints from models without imposing any assumptions about the model. Declare models impose constraints instead of explicitly stating event orders. Constraints can impose various choices and ordering of events; constraints typically have understandable names, but for details, refer to [5]. Current state-of-the-art mining tends to fail due to a computational explosion, and employ filtering to reduce this. Our tool is not intended to provide user-readable models, but instead to provide all constraints satisfied by a model. This allows post-processing to weed out uninteresting constraints, potentially obtaining better resulting models than making filtering beforehand out of necessity. Any post-processing (and complexity-reducing filtering) possible with existing miners is also possible with the UnconstrainedMiner; our miner just allows more intelligent post-processing due to having more information available, such as interactive filtering of models. In our demonstration, we show how the new miner can handle large event logs in short time, and how the resulting output can be imported into Excel for further processing. Our intended audience is researchers interested in Declare mining and users interested in abstract characterization of relationships between events. We explicitly do not target end-users who wish to see a Declare model for a particular log (but we are happy to demonstrate the miner on other concrete data).