22 Jun
2009
22 Jun
'09
10:05 a.m.
Johan Tibell
Typo? Bloom filters have O(1) lookup and tries O(m) lookup where m is the number of characters in the string.
Typically you need to examine the (whole) search string in order to compute the hash function, so I think it is fair to consider them both O(m). (Sorry about the alphabet confusion, I should of course have made it clear that I referred to search pattern size, not the data size.) -k -- If I haven't seen further, it is by standing in the footprints of giants