Djamal Belazzougui, Travis Gagie, Simon Gog, Giovanni Manzini, Jouni Sirén

Intuitively, if two strings $S_1$ and $S_2$ are sufficiently similar and we already have an FM-index for $S_1$ then, by storing a little extra information, we should be able to reuse parts of that index in an FM-index for $S_2$. We formalize this intuition and show that it can lead to significant space savings in practice, as well as to some interesting theoretical problems.

Knowledge Graph



Sign up or login to leave a comment