Monday, July 27, 2009

JANE: Journal and Author Name Estimator
I've just read over at Scienceroll about two tools that “help you determine which journal you should choose for publication”, based on your abstract or keywords [See Journal and Author Name Estimator and Huge Steps in Changing Science].

The first one is part of ResearchGATE (a “scientific network that connects researchers”) and the second one (which has been around for some time, since early 2008) comes as a standalone web application called "Journal and Author Name Estimator" (JANE). Note that to use the tool at ResearchGATE you must sign in.

Browsing around its site, I noticed that JANE was described in an article published in Bioinformatics in 20081, so I read the article and gave it a little try.

Here’s JANE’s description from its website:
Have you recently written a paper, but you're not sure to which journal you should submit it? Or maybe you want to find relevant articles to cite in your paper? Or are you an editor, and do you need to find reviewers for a particular paper? Jane can help!
Just enter the title and/or abstract of the paper in the box, and click on 'Find journals', 'Find authors' or 'Find Articles'. Jane will then compare your document to millions of documents in Medline to find the best matching journals, authors or articles
Also, here’s how it works (again, according to the info on its website -more details in the article-):
How does Jane work?
Jane first searches for the 50 articles that are most similar to your input*. For each of these articles, a similarity score between that article and your input is calculated. The similarity scores of all the articles belonging to a certain journal or author are summed to calculate the confidence score for that journal or author. The results are ranked by confidence score.
* For the computer geeks: we use the open source search engine Lucene. Queries using keywords are parsed with the QueryParser class, titles and abstracts are parsed using the MoreLikeThis parser class.
It’s hard for me to picture a submitting scientist SO lost that he/she  has absolutely no idea which journal publishes articles related to the work he/she is doing and with no clue whatsoever where to submit his/her work, to the extent that he/she would have to rely on a web-based application to enlighten him/her. I mean, surely he/she must have read an article or two for his/her own research and then have some sense of where he/she could submit. I agree that selecting a particular journal to submit your work is not a decision to be taken lightly, and in our case, after putting some thought into it, we generally narrow it down to 2-3 journals, but surely we know what journals have published work related (even remotely) to our own research in the past, from which we can pick our top 3.
In my opinion, using JANE for this is definitely not the way to go. I think it’s wiser to talk to colleagues at your department or at conferences. Their input will be more valuable and practical; some may have even submitted to some of the journals in your area and know a little about the inner workings of that particular journal. Also, at some conferences you can meet with journal’s representatives and discuss if your work would be suitable for their journal, or if you are lucky, they will approach you!
I tried JANE with several abstracts as query and it was kinda hit or miss: it generally gave the correct journal where the manuscript was originally published (not surprisingly, though) but some of the other journals in the list of “recommended journals” were not particularly related to the query as a whole, but to some particular words. This led JANE to propose journals (and this proposal was based on certain articles supposedly related to the query, published in those journals) that were way off.
Nevertheless, you will generally find journals that are indeed related to your input. The good thing is that you can check which articles JANE used to recommend the journals listed, so you can see if they are related as a whole or just because of a particular keyword in the abstract used as a query.
Furthermore, I fail to see an improvement large enough to justify not using directly Pubmed for this. If you put your keywords (although not whole abstracts) into Pubmed, you’ll get a series of articles with their respective journals and also, related papers. The way the authors justify using JANE over Pubmed is in no way satisfactory.
Despite all this, I think this tool can be useful for two things.
The first one is to discover “new” journals or to become aware of other journals (not from your typical selection) that have published related work in the past, so you can have a larger pool where to choose from. Also, you can narrow down your search to find Open Access journals, which is always good.
The second one is directed to editors or grant-awarding institutions: due to its “Find Authors” tool, it can help in finding reviewers for a particular article or grant.
Nevertheless, journals typically have a database with reviewers they generally use, but maybe a new journal (or editor) can benefit from this. I tried it with several articles in the different research areas I’ve been involved, and it generally turned up with a list of relevant researchers who would be appropriate for reviewing them. In my opinion, this is the best use for this tool.
In conclusion, I think that using this tool for “finding the most appropriate journal for publishing your results1 is definitely not the way to go, but I see some potential in using it to find related alternatives and, more importantly, as a tool for editors to find potential reviewers.

1Schuemie, M., & Kors, J. (2008). Jane: suggesting journals, finding experts Bioinformatics, 24 (5), 727-728 DOI: 10.1093/bioinformatics/btn006

ScienceBlips: vote it up!