D?j? Vu all over again

Added by Branwen Hide on 14 August 2009 14:55

Back in December of last year, I posted a blog about an online register developed by a group of French research students, Scientific Red Cards which highlights scientific papers that have found to be fraudulent or contain other types of academic misconduct.

The site seems to be going well; updated with news stories and comments on a semi-regularly bases, though the last entry into the database itself seems to be in 2008. I would love to think that this is because cases of research/academic misconduct are so few and far between, but I am not so sure on that. Especially since I have now come across another database, D?j? Vu, which highlights cases of similar citations in the scientific literature.

D?j? Vu is funded by the Hudson Foundation and the National Institutes of Health in the US, and seems to have been around for a while, having previously been picked up by Nature News, Nature Medicine, Science and Nucleic Acid Research. The database contains computationally identified and manually confirmed highly similar Medline citation pairs. The database also contains user-provided commentary and supporting information to substantiate the classification. The database is freely available to all user groups, and can be searched online or downloaded. They are also making available eTBLAST, a search tool which enables journal article submissions to be compared to an existing database to identify potential duplications before they are published.

However, despite how good the database and software sound, the site is not without controversy. The article in Science points out the Déj? vu casts a wide net, including ?innocent? papers (such as translations) along with suspicious ones. The database also challenges the norms of scientific publishing; gray areas such as review articles and when one copies previous introductions they have previously wrote ? is this really wrong? It will be interesting to see the direction the debate continues and how successful it becomes.

© Research Information Network 2005–2009