I understand that the PROV standard is to have arrows point up, as the edges mean wasGeneratedBy, etc. However, our experience is that having arrows go in that direction always requires explanation to scientists. Everytime that we showed the system to scientists when we had upward arrows, we needed to expalin this, always the first question. I think this is one of these issues where computer scientists are content to think about these arrows in either direction, but scientists find downward arrows completely natural and upward arrows confusing.
I understand that the PROV standard is to have arrows point up, as the edges mean wasGeneratedBy, etc. However, our experience is that having arrows go in that direction always requires explanation to scientists. Everytime that we showed the system to scientists when we had upward arrows, we needed to expalin this, always the first question. I think this is one of these issues where computer scientists are content to think about these arrows in either direction, but scientists find downward arrows completely natural and upward arrows confusing.