These Elsevier collaborations use machine learning to turn data into knowledge

Description

If there’s one thing scientists don’t lack these days, it’s information. Making sense of that information in a way that is useful, however, requires time and effort. From drug interaction information, to chemical reaction steps, to data on protein folding, a vast array of useful information is currently stored in formats that are designed to be read by humans, such as research articles, but cannot be easily leveraged computationally for other purposes.

Until recently, for scientific data to become readily accessible by other means, a human had to interpret documents and enter relevant data into a database by hand. And while humans do an excellent job of comprehending and synthesizing information, human effort doesn’t scale up the way computing power does. With the number and density of data sources ever on the rise, Elsevier and its collaborators hope to change the paradigm to enable computers to manage as much of the job of synthesizing data as possible, allowing humans to do better science with the results.