Saturday, February 13, 2010

Happy Valentine's Day, Operations Research, and Optimizing the Marriage Market

In an article in the European Journal of Operational Research (EJOR), entitled, "Optimizing the Marriage Market: An Application of the Linear Assignment Model," researchers based in Switzerland and England have developed a mathematical model whose solution yields a new social optimum as to who should be married to whom based on a longitudinal dataset in Switzerland.

I thought that this article (and associated research) was provocative and fitting given that we are celebrating Valentine's Day on February 14. The article appears in the April 2010 issue of EJOR, volume 202, pages 547-553. I read the article and thought that it was very interesting how the authors identified the appropriate weights to assign and how they assumed a central agency that would reassign the couples so as to achieve a better optimum. Interesting that, according to their study, they found that the existing marriage pairings using the dataset in Switzerland was very fair from optimal. The objective functions that the authors considered corresponded to age, education, ethnicity, and prior divorce history.

The media coverage surrounding the publication of this article has also been fascinating. See here for an article from the British Daily Mail.

I teach the assignment model and algorithm when I teach Management Science since these are classic operations research type problems. It is interesting to see extensions of the basic model developed and fitted to sociological longitudinal data.

Their formalism, I suspect, can update the methodological tools used by such companies and sites as match.com and eharmony.com

It's never dull to be keeping up with the scientific literature and, actually, it can be a lot of fun, as well, in addition to thought-provoking. There are not many existing societies in which such a top-down, central allocation, system-optimizing mechanism would be favorably accepted by the citizens, however!