Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

If your interview question's success or failure is highly correlated with the person's ability to apply a fairly esoteric data structure it's probably a bad question.

The person who has never come across Bloom filters before has a huge disadvantage in your question, and the person who has seen them before has a huge advantage. I'd guess that as soon as a candidate says "Bloom filter" before you do in an interview, it's an instant "hire" for you, but this is a bad sign since you will get lots of false negatives and false positives.

False negatives will be good devs who never needed a Bloom filter (there are many) and false positives will be green devs that have read the wikipedia article on Bloom filters just before the interview out of dumb luck, or had a question where that data structure was introduced in the way you're introducing it, etc.



So you're saying this interview question is a probabilistic (with the potential for false positives) tool for detecting membership in the set of people who are equipped to understand the concept? Sounds like a Bloom Filter to me.


Nah - if it's obvious the person has heard of a Bloom Filter before it's not a fair question to ask them but I will talk to them about the relationship between the size of the bit array and the number of hash functions to measure how deep the understanding goes.

I actually prefer it when people have never heard of a Bloom filter so that we can actually have a real discussion.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: