DISCOVERING UNEXPECTED
USES OF HASHES
Apart from the algorithms detailed in this book, other important algorithms are based
on hashes. For instance, the Locality-sensitive Hashing (LSH) algorithm relies on a large
number of hash functions to stitch apparently separated information together. If you
wonder how marketing companies and intelligence services put different chunks of
information together based on names and addresses that aren’t identical (for example,
guessing that “Los Angels,” “Los Angles,” and “Los Angleles” all refer to Los Angeles) the
answer is LSH. LSH chunks the information to check into parts and digests it using many
hash functions, resulting in the production of a special hash result, which is an address
for a bucket used to hold similar words. LSH is quite complex in its implementation, but
check out this material from the Massachusetts Institute of Technology (MIT):
http://
www.mit.edu/~andoni/LSH/
.
152
PART 2
Do'stlaringiz bilan baham: |