Elsevier wins CLEF SimpleText Track Task 1 (retrieve relevant passages given a scientific topic)



We have won the CLEF SimpleText Track, Task 1! CLEF is one of the most important annual conferences that hosts and organizes a number of challenges in the fields of Natural Language Processing and Text Mining. One of the challenges this year, was SimpleText (automatic simplification of scientific texts), and the key task (Task 1) has been to select relevant passages given a scientific topic, so that they can be included in a simplified summary of the topic.

Artemis and Hosein used some of the latest AI algorithms and participated with several runs, and as you will see from the rankings below, a couple of them made it at the top of the scoreboard!

Besides the fact that this is a great achievement on its own accord, research-wise, this work is super relevant to our Topic Pages work streams for improving the quality and enhancing the pages with new features. We are already utilizing this work in our PoCs, and I am hoping that the below provides also additional confidence to our stakeholders that we are delivering State-Of-the-Art results to our product.