How does stackexchange search for non English words work (or Korean specifically if that matters)?

Question

NOTE: originally attempted to ask on stackexchage meta, but couldn't due to some problems.

Let's take searching for this question as an example

If I search for "금수회이록", 금수회이록, 회이록, I get a 0 results page.

It's not like the search is ignoring non English characters either, because searching for something generic like 나는 gives a lot of results

I know that one of the answers contains the phrase minutes from a meeting of animals, so searching for minutes from a or meeting of animals brings up the post (although searching **minute** from a doesn't bring it up, so it seems the search engine cannot make a relationship between singular and plural of the same English word, but I suppose that is a problem for a different question)

So what is going on regarding non English character searches? Does the search results improve based on more number of people searching for the same stuff? Is the search just limited this way, when it comes to non English characters? By the way, searching on google for 금수회의록 stackexchange brings out this post as the first result.

Answer 1

Well, the linked question is about 금수회의록.

You searched for 금수회이록 - so it doesn't match. Search with the correct spelling and it matches.

That said, I don't see anything if I search for 회의록 (i.e., it doesn't understand 금수회의록 = 금수 + 회의록), so I think the segmentation algorithm could use some improvement...