In this gallery we can see the summaries of the clean_data dataset. We can see that Weak summary is clearly the most compact among all, but it may not convey the intuition. Slighlty bigger Strong summary can be compared with typed summaries even though it was computed taking into account only the data triples. In this case TypedWeak and TypedStrong summaries are identical. Bisimulation-based summaries fail to give any picture of the data.
