University of Michigan Uses AI to Mine Ecological Data From Online Comments

Researchers at the University of Michigan have developed a new artificial intelligence approach using tools like Claude and ChatGPT to extract and categorize rare ecological data from casual comments left by nature lovers on free online platforms such as eBird and iNaturalist.

Millions of people worldwide capture valuable data that can help researchers assess trends in ecosystem health without realizing it, according to a study led by the University of Michigan. Those observations remain hidden in an ocean of comments on platforms utilized by birdwatchers and nature enthusiasts.

Artificial Intelligence Mines Observations From Platforms Like eBird and iNaturalist

The study demonstrated that artificial intelligence tools can identify comments containing pertinent data, then extract and categorize that information to build structured datasets. According to Hengxing Zou—the study’s lead author who completed the research as a postdoctoral scholar at the U-M Institute for Global Change Biology—the possibilities here are considerable. The approach is really versatile and we have a lot of data out there basically waiting to be mined, Zou said.

Working alongside undergraduate researchers from ecology and computer science, the team crafted prompts that enabled Claude and ChatGPT to extract and classify interactions with good accuracy and precision. ChatGPT first isolated eBird data comments containing interactions. Olivia Stein, an undergraduate research assistant, then took a random sample of nearly 500 of those comments to classify the interaction types before having Claude perform the same task.

Direct Observation of Species Interactions Presents Challenges

While technological advances provide researchers with richer data on individuals and groups of single species in the wild, interactions between different species present a much greater hurdle. Direct observation is frequently required, yet maintaining field personnel to monitor the same patch of an ecosystem around the clock is nearly impossible.

In many cases, species interactions require direct observation and that’s really hard to come by simply because you can’t keep a lot of people in the field monitoring the same patch of an ecosystem for 24 hours a day, seven days a week, Zou explained. But it’s very important data to have. It’s connected to ecosystem function, ecosystem stability and a lot of the things we care about, especially in a changing global environment.

Proof-of-Concept Study Discovers Species Along America’s West Coast

During this proof-of-concept investigation, the research team found evidence of non-native plant species creeping into ecosystems along America’s West Coast that might otherwise have been missed. The study specifically focused on isolating comments capturing interactions between different species, such as a mixed-flock of a single snow goose feeding alongside Canada geese, or a West Coast lady butterfly resting on a bristly oxtongue flower in Sonoma County, California.

This approach gives us something ecologists have long needed: a way to examine species interactions across broad spatial and temporal scales, remarked Kai Zhu, who serves as a senior author and associate professor within the U-M School for Environment and Sustainability. AI can help turn scattered observations into ecological datasets that reveal how biological communities are responding to global change.

Funding Sources and Institutions Support the Research

The study is published in the Proceedings of the National Academy of Sciences. Financial support came from federal funding provided by the U.S. National Science Foundation and the U.S. Department of Agriculture. Additional funding sources included the David and Lucile Packard Foundation, the Alfred P. Sloan Foundation, the Institute for Global Change Biology at SEAS, the U-M Undergraduate Research Opportunity Program, and Michigan State University’s Ecology, Evolution, and Behavior Program, with computational resources supplied by Advanced Research Computing at U-M.

Other groups taking part in the study were organizations like the University of California Santa Cruz, University of California Davis, Georgia Institute of Technology, American Bird Conservancy, Notre Dame University, Santa Fe Institute, and Michigan State University.