I told a client today about a phenomenon known as “adversarial indexing.” Actually, I said it was “known as adversarial indexing” or that “information retrieval scientists call it adversarial indexing,” when I should have said “information retrieval scientists call it something like adversarial indexing.”
It refers to the concept of attempting to design systems that will serve relevant search results in a world governed by competitive rather than congruent motives: in a world where there is an incentive to spam the engines. A recent paper by some Microsoft scientists alluded to this, but I’m guessing they must have called it “adversarial information retrieval,” given that there are (or were, until this post) zero occurrences of the phrase “adversarial indexing” in the Google index if you do a search (using quotation marks).
In any case – adversarial indexing is a decent way to put it, because even deciding what gets into the index at all is dictated by the types of submitters you have. If many are trying to game the engine, you don’t even let some of them submit stuff, or you sandbox their stuff, or whatever. That’s not exactly an environment that has a long history.
Possibly the first appearance of this phrase in print was on the napkin my client wrote it on.
Shorthand for budding feature writers: “it’s like a game of cops and robbers!”
In 1999 Andrew co-founded Traffick.com, an acclaimed “guide to portals” which foresaw the rise of trends such as paid search and semantic analysis.