K-independent hashing

In computer science, a family of hash functions is said to be k-independent or k-universal if selecting a function at random from the family guarantees that the hash codes of any designated k keys are independent random variables