search

carnival-data / carnival

JVM property graph data unification framework
https://carnival-data.github.io/carnival/
GNU General Public License v3.0
7 stars 2 forks source link

JOSS paper - reasoning #111

Closed augustearth closed 3 months ago

augustearth commented 1 year ago

this JOSS paper has very little, almost nothing about reasoning I think, except by "reasons over the data to create case and control patient cohorts"; whereas the other Carnival paper (doi:10.3233/SHTI190178) talks more about the reasoners; the project website also includes reasoners in the Algorithms section https://carnival-data.github.io/carnival/#features; any reason why that's not explained in the JOSS paper as in the website and other paper?

augustearth commented 1 year ago

@hjwilli I am not sure we should claim that the demo does reasoning. Reasoning is the application of reason to arrive at a conclusion. While the demo does do some sorta complicated searching via the graph and produces results, it does not take the next step of drawing a conclusion. We could enhance it to do so... The conclusions could be "should be a case" and "should be a control" or something like that.

Alternatively, we could remove the word "reason" from the demo. We could add a use case for reasoning that talks about how GraphMethods are intended to encapsulated and enable reasoning. If we want to have an example, the Monday thing I wrote sorta does reasoning... It uses graph logic to make conclusions about the phase of a project based on the tasks, whether they are open or closed, whether they are assigned, and the dates of creation of the project board and tasks. It's not biomedical, but it does take the "conclusionating" step.

Thoughts?

kinow commented 1 year ago

@hjwilli any plans to tackle this one for the JOSS paper too? I cannot find a point in the checklist for review where this one would fit, but I do see a point of expanding on this functionality in the paper, or omitting it for simplicity.

hjwilli commented 3 months ago

Hi @kinow, I've updated the paper to remove the word reasoning and more specifically describe the graph algorithm carnival feature that is being used when describing the demonstration project.

kinow commented 3 months ago

Thanks @hjwilli !I left one comment in the commit, about an extra "with" in the text, I think. But I see no other blockers from my part. Take a look at my comment to see if you need to fix it in the text. Good luck with the rest of the review.

hjwilli commented 3 months ago

Thank you for the helpful review and for noticing the typo! It has been fixed.