With the stunning amount of 2.5 million scientific papers being published annually, there is no wonder why it is impossible for doctors to get through all of the material. Today, sites like MedCalc and UptoDate are useful tools for doctors in order to consult diagnostic criteria and double-checking treatment guidelines, Wired writes.
However, there is a lot of room for improvement and it is questioned whether artificial intelligence could be a solution to this science overload. This would come in the shape of machine learning assistants that could read incoming papers, refine its information and highlight any relevant findings.
This has resulted in Iris, the first attempt to create this type of AI. “The machine can currently read the abstract of a paper, map out its key concepts, and find papers relevant to those concepts. It provides a quick way to get a sense of the scientific landscape for a given topic, something especially useful when you don’t know the exact keywords for the type of research you are looking for”, Wired writes.
Alongside the release from Iris, the Allen Institute for Artificial Intelligence have recently launched their search engine, Semantic Scholar, taking research beyond keywords.
“One of the problems is getting research out of the dusty digital drawers and into the hands of people who can implement it,” Anita Schjøll Brede, the CEO of Iris told Wired. By the help of her tool it should be easier to navigate the literature for people conducting interdisciplinary research in particular.
Their three-year-plan is to develop a proactive version of the software that will remember the papers you read last week, while also providing you with new papers based on your project description. Within ten years, Schjøll Brede hopes that the AI will be powerful enough to discover new concepts solely based on its reading and understanding of the literature.
The machine powering Iris is “discipline-agnostic”, which means that it does not care if you wish to conduct research about cancer or composite material – it will research what you tell it to do, basically.
This article was first published at: https://www.wired.com/2016/11/artificial-intelligence-dig-cures-buried-online/