hiho@ll

think about this:
i have an ip (example: "192.168.0.1") and a name of a real existing person (example: "John Doe")

in my database i have names and ip addresses

now i need a algorithm which is very fast, which can calculate the difference between what is in the database and what input data i have

what is this all about?
i need a algorithm which can evaluate the difference in strings depending on its contents

for example:
if the user enters "Jon Doe" then my algorithm should find "John Doe" if there is no "Jon Doe" (seems it was a typo by the user (so the algorithm needs fuzzy logic))

why i need a hashing algorithm?
1. i need some scoring stuff which calculates a number which describes how much nonsense the user sent to me (not only yes = nonsense, no = ok (0,1) i also need: a bit nonsense, seems most right) and
2. i think the fastest method doing such stuff will be try to compare hashes

so my questions:
1. anybody has an idea of an hash algorithm which can replace the input done by the user?
(cause i thought, a string doesn't really differ to a hash, both are unique represenations, and hashes can/should be a number so it should be easier and faster to compare)
2. any idea on how i can do a fuzzy logic algorithm on hashes, which can manage the typo described above?

thx@ll

double post