A couple weeks ago, I came across the phrase assortative mating in an article about autism:

Judith Warner explores a provocative theory about why rates of autism, particularly the mild form known as Asperger's, are on the rise: because people who have certain “autistic” traits are increasingly meeting and marrying each other and having offspring who are more likely to be on the spectrum.

The theory of “assortative mating” was first put forth by neuroscientist Simon Baron-Cohen, a leading autism researcher[. . . .]

I hadn't seen the word “assortative” before, but it didn't occur to me to post about it here until a week or so later, when I came across it again in various articles about research on social networks in mythology:

The three myths were shown to be similar to real-life networks as they had similar degree distributions, were assortative and vulnerable to targeted attack. Assortativity is the tendency of a character of a certain degree to interact with a character of similar popularity; being vulnerable to targeted attack means that if you remove one of the most popular characters, it leads to a breakdown of the whole network—neither of these appears to happen in fiction.

Wikipedia has more general info on assortative mixing in the network-theory context, also known as assortativity, or (when referring specifically to social networks) as homophily.

According to MW11, the word “assortative” in the mating context (“being nonrandom mating based on like or unlike characteristics”) dates back to 1897; they don't list the network-theory meaning per se, but I can see how the one could have derived from the other.

(Btw, thanks to Google Web History search for letting me find the first article quickly and easily when I went looking for it after encountering the term a second time.)