DocumentCode :
2207764
Title :
Improved Consistent Sampling, Weighted Minhash and L1 Sketching
Author :
Ioffe, Sergey
Author_Institution :
Google Inc., Mountain View, CA, USA
fYear :
2010
fDate :
13-17 Dec. 2010
Firstpage :
246
Lastpage :
255
Abstract :
We propose a new Consistent Weighted Sampling method, where the probability of drawing identical samples for a pair of inputs is equal to their Jaccard similarity. Our method takes deterministic constant time per non-zero weight, improving on the best previous approach which takes expected constant time. The samples can be used as Weighted Minhash for efficient retrieval and compression (sketching) under Jaccard or L1 metric. A method is presented for using simple data statistics to reduce the running time of hash computation by two orders of magnitude. We compare our method with the random projection method and show that it has better characteristics for retrieval under L1. We present a novel method of mapping hashes to short bit-strings, apply it to Weighted Minhash, and achieve more accurate distance estimates from sketches than existing methods, as long as the inputs are sufficiently distinct. We show how to choose the optimal number of bits per hash for sketching, and demonstrate experimental results which agree with the theoretical analysis.
Keywords :
cryptography; data compression; file organisation; information retrieval; pattern matching; sampling methods; Jaccard similarity; L1 sketching; bit string; consistent weighted sampling method; data statistics; deterministic constant time; hash computation; random projection method; running time reduction; sample compression; sample retrieval; weighted Minhash; Compression; Hashing; Minhash; Retrieval; Sampling; Sketching;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Data Mining (ICDM), 2010 IEEE 10th International Conference on
Conference_Location :
Sydney, NSW
ISSN :
1550-4786
Print_ISBN :
978-1-4244-9131-5
Electronic_ISBN :
1550-4786
Type :
conf
DOI :
10.1109/ICDM.2010.80
Filename :
5693978
Link To Document :
بازگشت