TY - GEN
T1 - Parallel external memory suffix sorting
AU - Kärkkäinen, Juha
AU - Kempa, Dominik
AU - Puglisi, Simon J.
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2015.
PY - 2015
Y1 - 2015
N2 - Suffix sorting (or suffix array construction) is one of the most important tasks in string processing, with dozens of applications, particularly in text indexing and data compression. Some of these applications require the suffix array to be built for large inputs that greatly exceed the size of RAM and so external memory must be used. However, existing approaches for external memory suffix sorting either use debilitatingly large amounts of disk space, or become too slow when the size of the input data is more than a few times bigger than the size of RAM. In this paper we address the latter problem via a non-trivial parallelization of computation. In our experiments, the resulting algorithm is much faster than the best prior external memory algorithms while using very little disk space in addition to what is needed for the input and output. On the way to this result we provide the current fastest (parallel) internal memory algorithm for suffix sorting, which is usually around twice as fast as previous methods, while using around one quarter of the working space.
AB - Suffix sorting (or suffix array construction) is one of the most important tasks in string processing, with dozens of applications, particularly in text indexing and data compression. Some of these applications require the suffix array to be built for large inputs that greatly exceed the size of RAM and so external memory must be used. However, existing approaches for external memory suffix sorting either use debilitatingly large amounts of disk space, or become too slow when the size of the input data is more than a few times bigger than the size of RAM. In this paper we address the latter problem via a non-trivial parallelization of computation. In our experiments, the resulting algorithm is much faster than the best prior external memory algorithms while using very little disk space in addition to what is needed for the input and output. On the way to this result we provide the current fastest (parallel) internal memory algorithm for suffix sorting, which is usually around twice as fast as previous methods, while using around one quarter of the working space.
UR - https://www.scopus.com/pages/publications/84948991920
U2 - 10.1007/978-3-319-19929-0_28
DO - 10.1007/978-3-319-19929-0_28
M3 - Conference contribution
AN - SCOPUS:84948991920
SN - 9783319199283
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 329
EP - 342
BT - Combinatorial Pattern Matching - 26th Annual Symposium, CPM 2015, Proceedings
A2 - Vaccaro, Ugo
A2 - Porat, Ely
A2 - Cicalese, Ferdinando
PB - Springer Verlag
T2 - 26th Annual Symposium on Combinatorial Pattern Matching, CPM 2015
Y2 - 29 June 2015 through 1 July 2015
ER -