Why does searching for "limerence" give me a bunch of hits on "limericks"?

https://boards.straightdope.com/search?q=Limerence

Does the search engine simply ignore the letters past #5 or something? Very strange. Trying another example…psychopath and psychological didn’t give me irrelevant results… “Supercalifragilisticexpialidocious” didn’t give me any hits on Superman or such.

Any clues?

Part way down the search list it hits the one use of the word that Google can find on https://boards.straightdope.com/.

I’m guessing that with only 1 hit, it is giving false positives thinking you meant limerick. Not a great design but I think that is the answer.

Adding intentionally misspelled keywords to a website is a common SEO tactic.

True. But that does not explain why Discourse’s internal search engine would chase sorta-similar words in Discourse’s posts.

Maybe the Discourse designer have the unfortunate notion that the way Google does searches is the correct way.

This sort of fuzzy match can be useful for being able to still give results when you misspell a word or type one variant (e.g. finding dog when you type dogs) Most search engines will cut it out if you put the search term in quotes.

Including this one. Quoting the search will cut out fuzzy matches. But requires that you get the spelling correct. Also requires that the post you’re searching for gets the spelling the same as your search.

Fuzzy matching usually allows some number of characters to be wrong. It could be a fixed number or it could be proportional to the length of the term. You can play around with it and see how far off you can be and still get a match.

I find this sort of thing highly insulting. It seems to be based on the assumption that you are too stupid to tell it to search for what you really want, so it adds all sorts of rubbish as alleged “results”.

It’s better than the alternative, which is that it has the information that you want but doesn’t return it because it was too exact in its search. It makes sense to have some sort of algorithm to detect similarly spelled words, as those are often related.

People generally prefer getting some information they didn’t want over missing some information they did want.

Too stupid? You never make typos? What about a proper name that you just heard that you don’t know how to spell?

If you search for something misspelled or highly obscure that’s close to something you are much more likely to be searching for, Google gives you the hits for the more likely thing, but at the very top of the results page gives you the option to search for the exact string that you typed. It makes perfect sense to me.

If I search for new yirk, it says:

Showing results for new york
Search instead for new yirk [with hyperlink]

In Google, enclosing in quotation marks appears to make it look for the entire phrase, but not the exact phrase.

Searching for new yirk or “new yirk” (with or without quotation marks) gives exactly the same result . I think with misspellings like that, the intention is that if you really want to find that you can use the hyperlink at the top of the results that I describe in my prior post.

Whereas if you search for all of this string it gives hits focused on the word string; and if you search for “all of this string” it gives hits containing the entire phrase.

In creating the engines for searching,
The writers are clumsy and lurching.
They all try their best
And in fairness, i guess,
It’s unkind when one is besmirchive.

These “clumsy and lurching” writers started a company that has a market cap of $1.2 trillion.