[ad_1]
The investigative committee receives opinions from many smaller groups and considers a large amount of information to develop research strategies. Although the academy will not release the final recommendations of the committee to NASA within a few weeks, the scientists would like to know which of their issues will be adopted and which will be ruled out.
“The ten-year survey really helped NASA decide how they will lead the future of human discovery in space, so it is very important for them to understand the situation,” said Brandt Robertson, Professor of Astronomy and Astrophysics at the University of California, Santa Cruz.
A group of researchers hopes to use artificial intelligence to simplify this process. Their proposals are not aimed at specific tasks or asking questions; instead, they say that their artificial intelligence can help scientists make difficult decisions about which other proposals to prioritize.
The idea is that by training artificial intelligence to discover fast-growing or declining research areas, the tool can make it easier for investigative committees and groups to decide what should be included in the list.
“What we want is a system to complete the large amount of work done in the 10-year survey, and let the scientists engaged in the 10-year survey do what they do best,” said Haresson, Retired senior scientist and lead author at NASA Goddard Space Flight Center Proposed.
Although the members of each committee are selected on the basis of expertise in their respective fields, each member cannot grasp the nuances of each scientific topic. According to the authors, the number of astrophysics publications is increasing by 5% every year. Anyone needs to deal with a lot.
This is where Thronson’s artificial intelligence comes in.
The construction only took more than a year, but in the end, Thronson’s team was able to train it based on more than 400,000 studies published in the first ten years of the Astro2010 survey. They can also teach AI to sift through thousands of abstracts to identify areas of low and high influence from two- and three-word topic phrases such as “planetary system” or “exoplanet”.
According to the researchers’ white paper, artificial intelligence has successfully “reversed” six popular research topics in the past 10 years, including the rapid rise of exoplanet research and galaxy observation.
“One of the challenging aspects of artificial intelligence is that they sometimes predict, propose, or analyze things that are completely surprising,” Thronson said. “We have seen a lot.”
Thronson and his collaborators believe that the steering committee should use their artificial intelligence to help review and summarize the large amount of text that the team must sift through, allowing human experts to make the final decision.
Their research is not the first to try to use artificial intelligence to analyze and shape the scientific literature.Other AIs are already in use Help scientists conduct peer review The work of their colleagues.
But are tasks as important and influential as the 10-year survey worthy of trust?
[ad_2]
Source link
