作者: Ronitt Rubinfeld , Bernard Chazelle , Joe Kilian , Ayellet Tal
关键词: Bloom filter 、 Encoding (memory) 、 Function (mathematics) 、 Mathematics 、 Filter (video) 、 Algorithm 、 Lookup table 、 Set (abstract data type) 、 Theoretical computer science 、 Data structure 、 Hash function
摘要: We introduce the Bloomier filter, a data structure for compactly encoding function with static support in order to approximate evaluation queries. Our construction generalizes classical Bloom an ingenious hashing scheme heavily used networks and databases, whose main attribute---space efficiency---is achieved at expense of tiny false-positive rate. Whereas filters can handle only set membership queries, our deal arbitrary functions. give several designs varying simplicity optimality, we provide lower bounds prove (near) optimality constructions.