pcompress icon indicating copy to clipboard operation
pcompress copied to clipboard

Published 20 hours ago • verified publisher icon moinakg

frequency based chunking

Open max-mapper opened this issue 9 years ago • 1 comments

hey, was just reading this paper http://www-users.cs.umn.edu/~lv/FBC.pdf and looked for implementations but didn't find any. it proposes some pretty interesting space saving capabilities that go beyond e.g. rabin fingerprinting.

was just curious if you've seen that paper and considered that approach

max-mapper avatar Sep 18 '15 02:09 max-mapper

I have looked at this. However, the approach I used provides for high accuracy (>95% efficiency of brute force) and extreme space savings. It is possible to achieve petascale matching with a RAM based high-level index.

moinakg avatar Oct 08 '15 15:10 moinakg