We run a small search engine and have recently been notified of a number of hate related links in our results that would upset a significant proportion of our users. Our first instinct is to summarily remove these results, but I'm concerned that this makes us little better than the oppressive regimes that censor the web. Where does one draw the line and how might one justify removing results that we deem offensive?
Personally, I want to say that instead of removing the results, just make sure the results of an offensive nature are at the very bottom of the results.
Either that or take after Google and give users the option.
A simple way around your dilemma would be to allow the searcher themselves to decide. Instead of removing the results, you could flag them as "potentially containing hate speech".
Upon performing a search that contains these results, you could prompt the searcher:
This would allow the searcher to pick whether or not those results conflict with their values, and act accordingly. You could even cache their choice for later, so the dialog doesn't become a nuisance.
You'll still be stuck in the position of classifying & identifying "objectionable material" of sorts, and may catch some flak for that alone, but you avoid the whole issue of restricting your search results and allow the user to choose what's best for themselves.