Martijn J. Schuemie and Jan A. Kors (Bioinformatics doi:10.1093/bioinformatics/btn006 ) have created a freely available web-based application that, on the basis of a sample text, can suggest “journals and experts who have published similar articles”. Their aim is to help scientists to determine which journal is most appropriate for publishing their results, and which other scientists can be called upon to review their work. The application is called Jane (for journal/author name estimator).
I inputted some sample text to Jane, and was told that the Saudi Medical Journal was my top choice. No disrespect to that journal, but I know (because I am a person and not a computer) that this journal would be inappropriate for my test sample in at least two ways.
I would not primarily recommend an automatic selector to authors trying to decide where to submit their articles. When someone is ready to submit a paper, she or he will have given talks about the work and circulated drafts for comments from others in the field. That is a good time to ask for suggestions and advice about journals in which to publish. The scientist is then well-advised to read the author guidance on a few journals’ websites, to find out about editorial scope, impact factor and so on.
I think it is possibly counter-productive to use this kind of text-based comparison system on its own for making decisions about journal submission. At Nature, for example, we are looking for novel results, not something similar to what we have just published. Other journals are the same – most of them are looking for distinctive articles, not incremental repeats.
Rather than relying on computer searches to choose where to submit, I highly recommend looking at our free Author and Reviewers’ website for writing and submission advice. From there one can go straight to a great set of articles written by professional journal editors about how, where and why to submit and publish at the free science-information website SciDev.Net.
In addition, scientists can upload a draft manuscript into a community preprint server, where others in the field can comment and suggest. (Nature Precedings is one such, which provides meta-features such as alerting people in the field when new preprints have been uploaded, but many others. ArXiv is another, for the physical sciences.)
I think it will be a sad day when science journals publish “articles selected for us by computer”.