KEY TAKEAWAYS

• The increasing trend of artificially boosting citation metrics is an illegitimate, yet lucrative business.
• Several research teams are striving to identify and prevent these fraudulent practices, but the methods used are becoming alarmingly sophisticated.

Fraudulent publication tactics, from fake data to paper mills, pose a significant threat to the integrity of academic research. Now, increasing rates of citation manipulation are the latest trend to spark concerns among researchers. In a recent Nature News article, Dalmeet Singh Chawla looks at the scale of the threat and efforts to expose unscrupulous practices.

Spot the red flags

Computer scientist Yasir Zaki is among those at the forefront of investigations. As he explained to Singh Chawla, there are some key warning signs to look out for if citation manipulation is suspected:

• a steep rise in citations shortly after publication
• citations deriving from limited sources
• a sudden, large increase in citations.

The scale of the problem

While in the past, citation manipulation was a more ‘low-tech’ practice, with ‘citation rings’ citing each other’s work, we are presented with a very different picture today. Zaki’s team ran an undercover operation exposing a black market industry that sells citations via paper mills. Further work by the group found that fake preprints were a key method used to artificially bolster citation counts. In another demonstration of how easy ‘citation gaming’ has become in the digital age, a different group were able to list papers on Google Scholar that had been ‘authored’ by a cat and then cite these in fake papers they posted on ResearchGate.

Solutions in sight?

While the scale of this industry is sobering, Singh Chawla shone a light on efforts to tackle the issue:

• A tool by Guillaume Cabanac (University of Toulouse) detects unusual phrasing indicative of fake research papers. Cabanac reports that many of these papers also contain suspicious citations.
• Cyril Labbé’s group (Grenoble Alpes University) is developing a tool to flag unusual citation patterns.
• Zaki’s team suggest a new metric (the citation-concentration index) that identifies authors with citations derived from limited sources.

However, as fraudsters come up with new and nuanced ways to ‘game’ the system, the scientific community must remain vigilant.

————————————————–