This may just be a poor summary of the paper, but it sounds very strange. First, it uses "machine learning / an algorithm" to estimate the number of citations.

This may just be a poor summary of the paper, but it sounds very strange. First, it uses "machine learning / an algorithm" to estimate the number of citations. That suggests the system is to some extent a black box, which makes it hard to take the results seriously. Alternatively, if it is not a black box but is a detailed algorithm, then those details matter, but are not provided. And the bigger question is - why is "machine learning / an algorithm" needed here, what does it do that normal statistical analysis tools cannot? That is, find the numbers, correct for various factors, calculate values, etc. Those tools are well-understood, and seem to be sufficient to answer the question proposed here, yet they weren't used, so there is a lot more to this story.