Traditional models of academic publishing have been under attack from a number of directions, with factors such as the decreasing cost of publication and dissemination leading to the proliferation of online journals and alternative publishing models. One such alternative, straddling the border between blog and refereed publication, is Not Relevant, a web site recently created by Ian Soboroff as a venue for publishing and discussing work related to information retrieval that might have been rejected by traditional publication venues.
The goal of Not Relevant is to provide a novel dissemination venue for research in information retrieval, particularly when that research does not fit well in existing channels. Not Relevant strives for open dissemination of research, to put that research into the wild quickly, and to foster open and public discussion of that research.
The rest of the details are covered in the editorial guidelines, but the gist is that authors submit their manuscripts to the site, the board decides if there is sufficient merit to warrant publication (and the site is not subject to the same space restrictions as a conference), and assigns one or two members to write commentary on the submission. The paper itself is published through arXiv.org with a Creative Commons Attribution license.
Last week, we published our first submission dealing with detecting spam in blog datasets (the removal of which improves search results considerably).The paper is accompanied by commentary from Mounia Lalmas, and an (ongoing) discussion of various aspects of the work is recorded in the comments.
We hope that others will follow suit and submit papers for preview and discussion. This forum is not meant to replace peer review, but rather to facilitate more rapid and less formal dissemination of ideas around information retrieval. With your help, we can make Not Relevant truly useful.