Imagine receiving an email, not from a friend or a colleague, but from an unknown "editor" offering to publish your groundbreaking research—if you fork over a hefty fee. This unsettling reality is becoming all too common for scientists, and a new AI tool from the University of Colorado Boulder is stepping in to combat this alarming trend.

Led by computer scientist Daniel Acuña, this innovative platform automatically identifies "questionable" scientific journals, a growing issue in the research community. Published on August 27 in the journal Science Advances, this study shines a light on the rampant problem of so-called predatory journals.

These journals often target desperate researchers, enticing them with promises of publication in exchange for financial gain—fees that can reach thousands of dollars. Acuña frequently finds himself inundated with spam messages from these dubious journals, which often masquerade as legitimate publishers.

“There has been a growing effort among scientists and organizations to vet these journals,” Acuña explains. “But it’s like whack-a-mole. You catch one, and another pops up with a new name.”

The AI tool is designed to sift through the murky waters of scientific publishing, evaluating journals based on criteria such as the presence of a credible editorial board and the quality of their online content. While Acuña emphasizes that this tool is not foolproof, it represents a significant stride towards tackling the credibility crisis in scientific publishing.

In a world where misinformation proliferates, ensuring that research is built on a solid foundation is crucial. As Acuña pointedly remarks, “In science, you don’t start from scratch. You build on the research of others. So if the foundation of that tower crumbles, then the entire thing collapses.”

The peer review process, designed to uphold the quality of scientific inquiry, has been increasingly jeopardized by these unscrupulous publishers. Since the term “predatory journals” was coined by librarian Jeffrey Beall in 2009, the issue has only worsened, particularly affecting researchers from developing nations.

“They will say, ‘If you pay $500 or $1,000, we will review your paper,’” Acuña explains. “In reality, they don’t provide any service.”

In response to this crisis, organizations like the Directory of Open Access Journals (DOAJ) have worked tirelessly to identify and flag suspicious journals. However, with the rapid proliferation of these predatory outlets, the human effort to keep up has become overwhelming.

That’s where Acuña’s AI tool comes in. Trained with data from the DOAJ, the AI combed through a staggering list of nearly 15,200 open-access journals, identifying over 1,400 as potentially problematic. Although the AI made some misjudgments—flagging about 350 legitimate journals—this still left a substantial number of journals needing further scrutiny.

Acuña envisions this AI tool as an assistive technology—something to pre-screen the vast array of journals before handing the final decision to human experts. “I think this should be used as a helper to prescreen large numbers of journals,” he states, reiterating the importance of expert intervention.

Unlike other AI systems, which often operate as a “black box,” Acuña’s team has made an effort to ensure transparency in their tool. By analyzing data patterns, they discovered that questionable journals often publish an unusually high volume of articles and frequently include authors with multiple affiliations or who cite their own work excessively.

While this new AI system isn't publicly available yet, the research team hopes to share it with universities and publishing companies soon. Acuña believes it could serve as a vital “firewall for science,” protecting the integrity of research and data. “We know that when a new smartphone comes out, its software will have flaws, and we expect bug fixes in the future. We should probably do the same with science,” he concludes.

With co-authors from institutions across the globe, including Han Zhuang and Lizheng Liang, this study marks a pivotal step in the fight against the spread of questionable scientific publishing.