Hundreds of manuscripts are submitted to journals every day and we knew that was a problem that’s only getting worse. There’s of course nothing wrong with science being rapidly produced but its evaluation by journal editors is a work of nightmare.
After a paper gets submitted to a journal, it gets either accepted or rejected. An editor spends a sufficient amount of time starring at it, having difficulties to understand what’s it about, is it novel, or is the journal going to benefit by accepting the paper.
What we developed consisted of a tool letting editors upload manuscripts and then compare them to a huge corpus of already published papers using pure AI magic. This way, we were able to help editors make decisions that would normally either take long time, or lead to hurried rejection which would clog other journals while new science continues to get submitted.
Concept Group Trends
This is one of the most notable functionalities that we developed for the tool. It was an idea made possible by the fact that our technology can extract the most important terms from a paper and categorise them as drugs, methods, and diseases. On top of that, it looks for term groups that have been trending in the past few years, which may give an indication of novelty, or a topic being more and more discussed – what journals are mostly interested to publish.
An example could be that an old disease is treated with a new drug, or a new method has been introduced using the same drug to treat an old disease.
To support discovering novelty in submitted manuscripts, we decided to show when some of the most important terms of a paper are drastically trending up which would mean that these terms have been discussed very often.
Furthermore, we could roughly show a prediction of a paper’s impact in the next 3 years. This was, of course, not 100% accurate but more of a guidance to editors supporting their investigation.
We worked on a number of other features, including Author Background, Journal Match, Reviewer Finder, Key Statements, Related Papers, Technical Checks, and more, some of which became separate products.