Cool tutorial, and (obviously) bad science.
There is no reason to believe a title+abstract are enough to predict the publication venue of a paper.
Everyone knows that's ~80% determined by the author list #badumtssshttps://twitter.com/fchollet/status/1258084927884558336 …
To the contrary, it seems obvious that the keywords found in your abstract and title should be predictive of acceptance, because acceptance is tied to 1) your topic, 2) the strength of your results, both of which are reflected in the abstract/title. Also, reviews are blind.
-
-
I would need a much larger dataset (also with more fields) to be convinced it's not picking up on artifacts. Also, some reviews are blind, but did you actually vet the dataset for the tutorial? The submission page of IOVS does not mention blind reviewing anywhere for instance.
-
That reminds me - I would be very eager to believe a model that can guess the author list of a paper based solely on the bibliography list. Should be a strong signal.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.