Status: dormant; 2008-2012
Samir Khuller and I have been looking into algorithms for assigning papers to reviewers, e.g., for a conference or a funding agency such as NSF. We have proposed an algorithm based on min-cost flow and implemented it in Haskell. We have worked with Michael Hicks to apply the tool to POPL 2012, with good results, as described in Dr. Hicks's report. (The technique used here is unrelated to my least-unpopularity matching criteria project.)
Here is a git repository containing the source code of the matching tool and a short paper about it. Or you can read the PDF version of the paper directly.
Note: As an exception to the general policy for this web site, I do not waive my copyright to the paper, in the interest of maintaining the possibility of transferring the copyright for publication later.