Optimizing trie-based spelling correction algorithms at Constructor.io

I&#x27;m curious, how do you guys estimate &quot;fitness&quot; of the search results you found, i.e. whether it was what user searched for or not.<p>Just looking whether user later clicked on a result, or continued to search?

Does this mean all prefixes have to be regenerated for every addition of a new item? Or can you invalidate the relevant prefixes?

Nice post. How do you customize the Damerau-Levenshtein algorithm? Did you write a new from scratch?

Cool) sounds as interesting approach. It is good to compare it in real life.