Automation is reaching into a vast range of professions, and my own is certainly no different. Through this blog and various other means, I try to locate interesting research and practices from around the world, and bring them together into some kind of narrative.
With so much going on around the world, it stands to reason that a computer could be trained to do a similar task, and that’s certainly the aim of Semantic Scholar, a new tool launched by The Allen Institute for Artificial Intelligence.
Automated horizon scanning
The tool offers users a means of hunting for papers in specific fields, and then filter your search by date, publication and so on. The developers believe that Semantic Scholar stands above the likes of Google Scholar due to the smarter way in which it hunts down relevant papers.
The software has been trained to better understand the nuances of particular papers using a range of machine learning techniques.
Of course, the project isn’t the only attempt to better empower algorithms to make sense of the patterns they find in data. A second service, called Meta, was also launched recently to try and ease the way in which we can identify and collaborate with people mentioned in medical literature.
The aim of the service is to build an understanding of a user based upon their previous reading habits, and thus recommend papers it thinks the user will enjoy.
“Essentially, it allows you to track at the concept level, or the technology level, rather than the article level,” the team say. “Concepts like the CRISPR technology, which is really revolutionizing how genome engineering is happening right now—we picked that up as an emerging concept a number of years ago.”
Of course, at the moment the platforms both provide wayfinding services, albeit rather smart ones. Neither seems to do much to help users make sense of the papers themselves, which given the inherent complexities of academic research still presents a hurdle to be overcome if greater university/business collaboration is to be achieved.
The team at Semantic Scholar do appear to be working on that however, with attempts to better understand things like graphs or charts within papers. The ultimate goal is to be able to quickly tell the user whether a paper is worthwhile or not.
If you have an interest in keeping on top of the research landscape, then both Semantic Scholar and Meta are projects to certainly keep an eye on.