Citizen science, supercomputers and AI Posted on : Jan 08 - 2022

Citizen scientists have helped researchers discover new types of galaxies, design drugs to fight COVID-19, and map the bird world. The term describes a range of ways that the public can meaningfully contribute to scientific and engineering research, as well as environmental monitoring.

As members of the Computing Community Consortium (CCC) recently argued in a Quadrennial Paper, "Imagine All the People: Citizen Science, Artificial Intelligence, and Computational Research," non-scientists can help advance science by "providing or analyzing data at spatial and temporal resolutions or scales and speeds that otherwise would be impossible given limited staff and resources."

Recently, citizen scientists' efforts have found a new purpose: helping researchers develop machine learning models, using labeled data and algorithms, to train a computer to solve a specific task.

This approach was pioneered by the crowdsourced astronomy project Galaxy Zoo, which started leveraging citizen scientists in 2007. In 2019, researchers used labeled data to train a neural network model to classify hundreds of millions of unlabeled galaxies.

"Using the millions of classifications carried out by the public in the Galaxy Zoo project to train a neural network is an inspiring use of the citizens science program," said Elise Jennings, a computer scientist at Argonne Leadership Computing Facility (ALCF) who contributed to the effort.

TACC is supporting a number of projects—from identifying fake news to pinpointing structures in danger during natural hazards—that use citizen science to train AI models and enable new scientific successes.

Tinder for galaxies

The Hobby-Eberly Telescope Dark Energy Experiment, or HETDEX, is the first major experiment to search for evolution in dark energy. Based at the McDonald Observatory in West Texas, it looks deeper into the past than ever before to determine with great accuracy how fast the universe is accelerating.

The experiment relies on being able to identify the location, distance, and redshift of tens of millions of galaxies. But Karl Gebhardt, a professor of Astronomy at The University of Texas at Austin (UT Austin) and lead scientist on the project, faced a problem. The computational algorithms were having difficulty separating real target galaxies from false positives.

Strangely enough, humans can detect the difference easily. So, working with graduate students Lindsay House and Dustin Davis, and data scientist Erin Mentuch Cooper, they created a citizen science app called 'Dark Energy Explorers' to train a machine learning algorithm to assist in the process.

Individuals with minimal training can look at spectral lines and images of point sources and swipe left or right, depending on whether they believe it is a real galaxy or something else such as an artifact of the algorithm or a speck of dust on the sensor. The app has jokingly been called "Tinder for Galaxies," Gebhardt says. To date, citizen scientists have made almost 2 million classifications and more are needed. View more