arXiv subjects

Tomasz Kociumaka

Publications and source records attributed to Tomasz Kociumaka.

At least 73 records · Page 4Linked to original sources

arXiv2021

Towards a Definitive Compressibility Measure for Repetitive Sequences

Unlike in statistical compression, where Shannon's entropy is a definitive lower bound, no such clear measure exists for the compressibility of repetitive sequences. Since statistical entropy does not capture repetitiveness, ad-hoc measures like the size $z$ of the Lempel--Ziv parse are frequently used to estimate it. The size $b \le z$ of the smallest bidirectional macro scheme captures better what can be achieved via copy-paste processes, though it is NP-complete to compute and it is not monotonic upon symbol appends. Recently, a more principled measure, the size $γ$ of the smallest string \emph{attractor}, was introduced. The measure $γ\le b$ lower bounds all the previous relevant ones, yet length-$n$ strings can be represented and efficiently indexed within space $O(γ\log\frac{n}γ)$, which also upper bounds most measures. While $γ$ is certainly a better measure of repetitiveness than $b$, it is also NP-complete to compute and not monotonic, and it is unknown if one can always represent a string in $o(γ\log n)$ space. In this paper, we study an even smaller measure, $δ\le γ$, which can be computed in linear time, is monotonic, and allows encoding every string in $O(δ\log\frac{n}δ)$ space because $z = O(δ\log\frac{n}δ)$. We show that $δ$ better captures the compressibility of repetitive strings. Concretely, we show that (1) $δ$ can be strictly smaller than $γ$, by up to a logarithmic factor; (2) there are string families needing $Ω(δ\log\frac{n}δ)$ space to be encoded, so this space is optimal for every $n$ and $δ$; (3) one can build run-length context-free grammars of size $O(δ\log\frac{n}δ)$, whereas the smallest (non-run-length) grammar can be up to $Θ(\log n/\log\log n)$ times larger; and (4) within $O(δ\log\frac{n}δ)$ space we can not only...

cs.DS↗

arXiv2020

Tight Bound for the Number of Distinct Palindromes in a Tree

For an undirected tree with $n$ edges labelled by single letters, we consider its substrings, which are labels of the simple paths between pairs of nodes. We prove that there are $O(n^{1.5})$ different palindromic substrings. This solves an open problem of Brlek, Lafrenière, and Provençal (DLT 2015), who gave a matching lower-bound construction. Hence, we settle the tight bound of $Θ(n^{1.5})$ for the maximum palindromic complexity of trees. For standard strings, i.e., for paths, the palindromic complexity is $n+1$. We also propose $O(n^{1.5} \log{n})$-time algorithm for reporting all distinct palindromes in an undirected tree with $n$ edges.

cs.DS↗

arXiv2020

Resolution of the Burrows-Wheeler Transform Conjecture

The Burrows-Wheeler Transform (BWT) is an invertible text transformation that permutes symbols of a text according to the lexicographical order of its suffixes. BWT is the main component of popular lossless compression programs (such as bzip2) as well as recent powerful compressed indexes (such as $r$-index [Gagie et al., J. ACM, 2020]), central in modern bioinformatics. The compression ratio of BWT is quantified by the number $r$ of equal-letter runs. Despite the practical significance of BWT, no non-trivial bound on the value of $r$ is known. This is in contrast to nearly all other known compression methods, whose sizes have been shown to be either always within a ${\rm polylog}\,n$ factor (where $n$ is the length of text) from $z$, the size of Lempel-Ziv (LZ77) parsing of the text, or significantly larger in the worst case (by a $n^{\varepsilon}$ factor for $\varepsilon > 0$). In this paper, we show that $r = \mathcal{O}(z \log^2n)$ holds for every text. This result has numerous implications for text indexing and data compression; for example: (1) it proves that many results related to BWT automatically apply to methods based on LZ77, e.g., it is possible to obtain functionality of the suffix tree in $\mathcal{O}(z\,{\rm polylog}\,n)$ space; (2) it shows that many text processing tasks can be solved in the optimal time assuming the text is compressible using LZ77 by a sufficiently large ${\rm polylog}\,n$ factor; (3) it implies the first non-trivial relation between the number of runs in the BWT of the text and its reverse. In addition, we provide an $\mathcal{O}(z\,{\rm polylog}\,n)$-time algorithm converting the LZ77 parsing into the run-length compressed BWT. To achieve this, we develop a number of new data structures and techniques of independent interest.

cs.DS↗

arXiv2020

Faster Approximate Pattern Matching: A Unified Approach

Approximate pattern matching is a natural and well-studied problem on strings: Given a text $T$, a pattern $P$, and a threshold $k$, find (the starting positions of) all substrings of $T$ that are at distance at most $k$ from $P$. We consider the two most fundamental string metrics: the Hamming distance and the edit distance. Under the Hamming distance, we search for substrings of $T$ that have at most $k$ mismatches with $P$, while under the edit distance, we search for substrings of $T$ that can be transformed to $P$ with at most $k$ edits. Exact occurrences of $P$ in $T$ have a very simple structure: If we assume for simplicity that $|T| \le 3|P|/2$ and trim $T$ so that $P$ occurs both as a prefix and as a suffix of $T$, then both $P$ and $T$ are periodic with a common period. However, an analogous characterization for the structure of occurrences with up to $k$ mismatches was proved only recently by Bringmann et al. [SODA'19]: Either there are $O(k^2)$ $k$-mismatch occurrences of $P$ in $T$, or both $P$ and $T$ are at Hamming distance $O(k)$ from strings with a common period $O(m/k)$. We tighten this characterization by showing that there are $O(k)$ $k$-mismatch occurrences in the case when the pattern is not (approximately) periodic, and we lift it to the edit distance setting, where we tightly bound the number of $k$-edit occurrences by $O(k^2)$ in the non-periodic case. Our proofs are constructive and let us obtain a unified framework for approximate pattern matching for both considered distances. We showcase the generality of our framework with results for the fully-compressed setting (where $T$ and $P$ are given as a straight-line program) and for the dynamic setting (where we extend a data structure of Gawrychowski et al. [SODA'18]).

cs.DS↗

arXiv2020

Sublinear-Time Algorithms for Computing & Embedding Gap Edit Distance

In this paper, we design new sublinear-time algorithms for solving the gap edit distance problem and for embedding edit distance to Hamming distance. For the gap edit distance problem, we give an $\tilde{O}(\frac{n}{k}+k^2)$-time greedy algorithm that distinguishes between length-$n$ input strings with edit distance at most $k$ and those with edit distance exceeding $(3k+5)k$. This is an improvement and a simplification upon the result of Goldenberg, Krauthgamer, and Saha [FOCS 2019], where the $k$ vs $Θ(k^2)$ gap edit distance problem is solved in $\tilde{O}(\frac{n}{k}+k^3)$ time. We further generalize our result to solve the $k$ vs $k'$ gap edit distance problem in time $\tilde{O}(\frac{nk}{k'}+k^2+ \frac{k^2}{k'}\sqrt{nk})$, strictly improving upon the previously known bound $\tilde{O}(\frac{nk}{k'}+k^3)$. Finally, we show that if the input strings do not have long highly periodic substrings, then already the $k$ vs $(1+ε)k$ gap edit distance problem can be solved in sublinear time. Specifically, if the strings contain no substring of length $\ell$ with period at most $2k$, then the running time we achieve is $\tilde{O}(\frac{n}{ε^2 k}+k^2\ell)$. We further give the first sublinear-time probabilistic embedding of edit distance to Hamming distance. For any parameter $p$, our $\tilde{O}(\frac{n}{p})$-time procedure yields an embedding with distortion $O(kp)$, where $k$ is the edit distance of the original strings. Specifically, the Hamming distance of the resultant strings is between $\frac{k-p+1}{p+1}$ and $O(k^2)$ with good probability. This generalizes the linear-time embedding of Chakraborty, Goldenberg, and Koucký [STOC 2016], where the resultant Hamming distance is between $\frac k2$ and $O(k^2)$. Our algorithm is based on a random walk over samples, which we believe will find other applications in sublinear-time algorithms.

cs.DS↗

arXiv2020

Improved Circular $k$-Mismatch Sketches

The shift distance $\mathsf{sh}(S_1,S_2)$ between two strings $S_1$ and $S_2$ of the same length is defined as the minimum Hamming distance between $S_1$ and any rotation (cyclic shift) of $S_2$. We study the problem of sketching the shift distance, which is the following communication complexity problem: Strings $S_1$ and $S_2$ of length $n$ are given to two identical players (encoders), who independently compute sketches (summaries) $\mathtt{sk}(S_1)$ and $\mathtt{sk}(S_2)$, respectively, so that upon receiving the two sketches, a third player (decoder) is able to compute (or approximate) $\mathsf{sh}(S_1,S_2)$ with high probability. This paper primarily focuses on the more general $k$-mismatch version of the problem, where the decoder is allowed to declare a failure if $\mathsf{sh}(S_1,S_2)>k$, where $k$ is a parameter known to all parties. Andoni et al. (STOC'13) introduced exact circular $k$-mismatch sketches of size $\widetilde{O}(k+D(n))$, where $D(n)$ is the number of divisors of $n$. Andoni et al. also showed that their sketch size is optimal in the class of linear homomorphic sketches. We circumvent this lower bound by designing a (non-linear) exact circular $k$-mismatch sketch of size $\widetilde{O}(k)$; this size matches communication-complexity lower bounds. We also design $(1\pm \varepsilon)$-approximate circular $k$-mismatch sketch of size $\widetilde{O}(\min(\varepsilon^{-2}\sqrt{k}, \varepsilon^{-1.5}\sqrt{n}))$, which improves upon an $\widetilde{O}(\varepsilon^{-2}\sqrt{n})$-size sketch of Crouch and McGregor (APPROX'11).

cs.DS↗

arXiv2020

Counting Distinct Patterns in Internal Dictionary Matching

We consider the problem of preprocessing a text $T$ of length $n$ and a dictionary $\mathcal{D}$ in order to be able to efficiently answer queries $CountDistinct(i,j)$, that is, given $i$ and $j$ return the number of patterns from $\mathcal{D}$ that occur in the fragment $T[i \mathinner{.\,.} j]$. The dictionary is internal in the sense that each pattern in $\mathcal{D}$ is given as a fragment of $T$. This way, the dictionary takes space proportional to the number of patterns $d=|\mathcal{D}|$ rather than their total length, which could be $Θ(n\cdot d)$. An $\tilde{\mathcal{O}}(n+d)$-size data structure that answers $CountDistinct(i,j)$ queries $\mathcal{O}(\log n)$-approximately in $\tilde{\mathcal{O}}(1)$ time was recently proposed in a work that introduced internal dictionary matching [ISAAC 2019]. Here we present an $\tilde{\mathcal{O}}(n+d)$-size data structure that answers $CountDistinct(i,j)$ queries $2$-approximately in $\tilde{\mathcal{O}}(1)$ time. Using range queries, for any $m$, we give an $\tilde{\mathcal{O}}(\min(nd/m,n^2/m^2)+d)$-size data structure that answers $CountDistinct(i,j)$ queries exactly in $\tilde{\mathcal{O}}(m)$ time. We also consider the special case when the dictionary consists of all square factors of the string. We design an $\mathcal{O}(n \log^2 n)$-size data structure that allows us to count distinct squares in a text fragment $T[i \mathinner{.\,.} j]$ in $\mathcal{O}(\log n)$ time.

cs.DS↗

arXiv2020

Efficient Representation and Counting of Antipower Factors in Words

A $k$-antipower (for $k \ge 2$) is a concatenation of $k$ pairwise distinct words of the same length. The study of fragments of a word being antipowers was initiated by Fici et al. (ICALP 2016) and first algorithms for computing such fragments were presented by Badkobeh et al. (Inf. Process. Lett., 2018). We address two open problems posed by Badkobeh et al. We propose efficient algorithms for counting and reporting fragments of a word which are $k$-antipowers. They work in $\mathcal{O}(nk \log k)$ time and $\mathcal{O}(nk \log k + C)$ time, respectively, where $C$ is the number of reported fragments. For $k=o(\sqrt{n/\log n})$, this improves the time complexity of $\mathcal{O}(n^2/k)$ of the solution by Badkobeh et al. We also show that the number of different $k$-antipower factors of a word of length $n$ can be computed in $\mathcal{O}(nk^4 \log k \log n)$ time. Our main algorithmic tools are runs and gapped repeats. Finally we present an improved data structure that checks, for a given fragment of a word and an integer $k$, if the fragment is a $k$-antipower. This is a full and extended version of a paper from LATA 2019. In particular, all results about counting different antipowers factors are completely new compared with the LATA proceedings version.

cs.DS↗

arXiv2020

Time-Space Tradeoffs for Finding a Long Common Substring

We consider the problem of finding, given two documents of total length $n$, a longest string occurring as a substring of both documents. This problem, known as the Longest Common Substring (LCS) problem, has a classic $O(n)$-time solution dating back to the discovery of suffix trees (Weiner, 1973) and their efficient construction for integer alphabets (Farach-Colton, 1997). However, these solutions require $Θ(n)$ space, which is prohibitive in many applications. To address this issue, Starikovskaya and Vildhøj (CPM 2013) showed that for $n^{2/3} \le s \le n^{1-o(1)}$, the LCS problem can be solved in $O(s)$ space and $O(\frac{n^2}{s})$ time. Kociumaka et al. (ESA 2014) generalized this tradeoff to $1 \leq s \leq n$, thus providing a smooth time-space tradeoff from constant to linear space. In this paper, we obtain a significant speed-up for instances where the length $L$ of the sought LCS is large. For $1 \leq s \leq n$, we show that the LCS problem can be solved in $O(s)$ space and $\tilde{O}(\frac{n^2}{L\cdot s}+n)$ time. The result is based on techniques originating from the LCS with Mismatches problem (Flouri et al., 2015; Charalampopoulos et al., CPM 2018), on space-efficient locally consistent parsing (Birenzwige et al., SODA 2020), and on the structure of maximal repetitions (runs) in the input documents.

cs.DS↗

arXiv2020

Approximating longest common substring with $k$ mismatches: Theory and practice

In the problem of the longest common substring with $k$ mismatches we are given two strings $X, Y$ and must find the maximal length $\ell$ such that there is a length-$\ell$ substring of $X$ and a length-$\ell$ substring of $Y$ that differ in at most $k$ positions. The length $\ell$ can be used as a robust measure of similarity between $X, Y$. In this work, we develop new approximation algorithms for computing $\ell$ that are significantly more efficient that previously known solutions from the theoretical point of view. Our approach is simple and practical, which we confirm via an experimental evaluation, and is probably close to optimal as we demonstrate via a conditional lower bound.

cs.DS↗

arXiv2020

The Streaming k-Mismatch Problem: Tradeoffs between Space and Total Time

We revisit the $k$-mismatch problem in the streaming model on a pattern of length $m$ and a streaming text of length $n$, both over a size-$σ$ alphabet. The current state-of-the-art algorithm for the streaming $k$-mismatch problem, by Clifford et al. [SODA 2019], uses $\tilde O(k)$ space and $\tilde O\big(\sqrt k\big)$ worst-case time per character. The space complexity is known to be (unconditionally) optimal, and the worst-case time per character matches a conditional lower bound. However, there is a gap between the total time cost of the algorithm, which is $\tilde O(n\sqrt k)$, and the fastest known offline algorithm, which costs $\tilde O\big(n + \min\big(\frac{nk}{\sqrt m},σn\big)\big)$ time. Moreover, it is not known whether improvements over the $\tilde O(n\sqrt k)$ total time are possible when using more than $O(k)$ space. We address these gaps by designing a randomized streaming algorithm for the $k$-mismatch problem that, given an integer parameter $k\le s \le m$, uses $\tilde O(s)$ space and costs $\tilde O\big(n+\min\big(\frac {nk^2}m,\frac{nk}{\sqrt s},\frac{σnm}s\big)\big)$ total time. For $s=m$, the total runtime becomes $\tilde O\big(n + \min\big(\frac{nk}{\sqrt m},σn\big)\big)$, which matches the time cost of the fastest offline algorithm. Moreover, the worst-case time cost per character is still $\tilde O\big(\sqrt k\big)$.

cs.DS↗

arXiv2020

Circular Pattern Matching with $k$ Mismatches

The $k$-mismatch problem consists in computing the Hamming distance between a pattern $P$ of length $m$ and every length-$m$ substring of a text $T$ of length $n$, if this distance is no more than $k$. In many real-world applications, any cyclic rotation of $P$ is a relevant pattern, and thus one is interested in computing the minimal distance of every length-$m$ substring of $T$ and any cyclic rotation of $P$. This is the circular pattern matching with $k$ mismatches ($k$-CPM) problem. A multitude of papers have been devoted to solving this problem but, to the best of our knowledge, only average-case upper bounds are known. In this paper, we present the first non-trivial worst-case upper bounds for the $k$-CPM problem. Specifically, we show an $O(nk)$-time algorithm and an $O(n+\frac{n}{m}\,k^4)$-time algorithm. The latter algorithm applies in an extended way a technique that was very recently developed for the $k$-mismatch problem [Bringmann et al., SODA 2019]. A preliminary version of this work appeared at FCT 2019. In this version we improve the time complexity of the main algorithm from $O(n+\frac{n}{m}\,k^5)$ to $O(n+\frac{n}{m}\,k^4)$.

cs.DS↗

arXiv2020

Approximating Text-to-Pattern Hamming Distances

We revisit a fundamental problem in string matching: given a pattern of length m and a text of length n, both over an alphabet of size $σ$, compute the Hamming distance between the pattern and the text at every location. Several $(1+ε)$-approximation algorithms have been proposed in the literature, with running time of the form $O(ε^{-O(1)}n\log n\log m)$, all using fast Fourier transform (FFT). We describe a simple $(1+ε)$-approximation algorithm that is faster and does not need FFT. Combining our approach with additional ideas leads to numerous new results: - We obtain the first linear-time approximation algorithm; the running time is $O(ε^{-2}n)$. - We obtain a faster exact algorithm computing all Hamming distances up to a given threshold k; its running time improves previous results by logarithmic factors and is linear if $k\le\sqrt m$. - We obtain approximation algorithms with better $ε$-dependence using rectangular matrix multiplication. The time-bound is $Õ(n)$ when the pattern is sufficiently long: $m\ge ε^{-28}$. Previous algorithms require $Õ(ε^{-1}n)$ time. - When k is not too small, we obtain a truly sublinear-time algorithm to find all locations with Hamming distance approximately (up to a constant factor) less than k, in $O((n/k^{Ω(1)}+occ)n^{o(1)})$ time, where occ is the output size. The algorithm leads to a property tester, returning true if an exact match exists and false if the Hamming distance is more than $δm$ at every location, running in $Õ(δ^{-1/3}n^{2/3}+δ^{-1}n/m)$ time. - We obtain a streaming algorithm to report all locations with Hamming distance approximately less than k, using $Õ(ε^{-2}\sqrt k)$ space. Previously, streaming algorithms were known for the exact problem with Õ(k) space or for the approximate problem with $Õ(ε^{-O(1)}\sqrt m)$ space.

cs.DS↗

arXiv2019

Weighted Shortest Common Supersequence Problem Revisited

A weighted string, also known as a position weight matrix, is a sequence of probability distributions over some alphabet. We revisit the Weighted Shortest Common Supersequence (WSCS) problem, introduced by Amir et al. [SPIRE 2011], that is, the SCS problem on weighted strings. In the WSCS problem, we are given two weighted strings $W_1$ and $W_2$ and a threshold $\mathit{Freq}$ on probability, and we are asked to compute the shortest (standard) string $S$ such that both $W_1$ and $W_2$ match subsequences of $S$ (not necessarily the same) with probability at least $\mathit{Freq}$. Amir et al. showed that this problem is NP-complete if the probabilities, including the threshold $\mathit{Freq}$, are represented by their logarithms (encoded in binary). We present an algorithm that solves the WSCS problem for two weighted strings of length $n$ over a constant-sized alphabet in $\mathcal{O}(n^2\sqrt{z} \log{z})$ time. Notably, our upper bound matches known conditional lower bounds stating that the WSCS problem cannot be solved in $\mathcal{O}(n^{2-\varepsilon})$ time or in $\mathcal{O}^*(z^{0.5-\varepsilon})$ time unless there is a breakthrough improving upon long-standing upper bounds for fundamental NP-hard problems (CNF-SAT and Subset Sum, respectively). We also discover a fundamental difference between the WSCS problem and the Weighted Longest Common Subsequence (WLCS) problem, introduced by Amir et al. [JDA 2010]. We show that the WLCS problem cannot be solved in $\mathcal{O}(n^{f(z)})$ time, for any function $f(z)$, unless $\mathrm{P}=\mathrm{NP}$.

cs.DS↗

arXiv2019

Internal Dictionary Matching

We introduce data structures answering queries concerning the occurrences of patterns from a given dictionary $\mathcal{D}$ in fragments of a given string $T$ of length $n$. The dictionary is internal in the sense that each pattern in $\mathcal{D}$ is given as a fragment of $T$. This way, $\mathcal{D}$ takes space proportional to the number of patterns $d=|\mathcal{D}|$ rather than their total length, which could be $Θ(n\cdot d)$. In particular, we consider the following types of queries: reporting and counting all occurrences of patterns from $\mathcal{D}$ in a fragment $T[i..j]$ and reporting distinct patterns from $\mathcal{D}$ that occur in $T[i..j]$. We show how to construct, in $\mathcal{O}((n+d) \log^{\mathcal{O}(1)} n)$ time, a data structure that answers each of these queries in time $\mathcal{O}(\log^{\mathcal{O}(1)} n+|output|)$. The case of counting patterns is much more involved and needs a combination of a locally consistent parsing with orthogonal range searching. Reporting distinct patterns, on the other hand, uses the structure of maximal repetitions in strings. Finally, we provide tight---up to subpolynomial factors---upper and lower bounds for the case of a dynamic dictionary.

cs.DS↗

arXiv2019

Optimal-Time Dictionary-Compressed Indexes

We describe the first self-indexes able to count and locate pattern occurrences in optimal time within a space bounded by the size of the most popular dictionary compressors. To achieve this result we combine several recent findings, including \emph{string attractors} --- new combinatorial objects encompassing most known compressibility measures for highly repetitive texts ---, and grammars based on \emph{locally-consistent parsing}. More in detail, let $γ$ be the size of the smallest attractor for a text $T$ of length $n$. The measure $γ$ is an (asymptotic) lower bound to the size of dictionary compressors based on Lempel--Ziv, context-free grammars, and many others. The smallest known text representations in terms of attractors use space $O(γ\log(n/γ))$, and our lightest indexes work within the same asymptotic space. Let $ε>0$ be a suitably small constant fixed at construction time, $m$ be the pattern length, and $occ$ be the number of its text occurrences. Our index counts pattern occurrences in $O(m+\log^{2+ε}n)$ time, and locates them in $O(m+(occ+1)\log^εn)$ time. These times already outperform those of most dictionary-compressed indexes, while obtaining the least asymptotic space for any index searching within $O((m+occ)\,\textrm{polylog}\,n)$ time. Further, by increasing the space to $O(γ\log(n/γ)\log^εn)$, we reduce the locating time to the optimal $O(m+occ)$, and within $O(γ\log(n/γ)\log n)$ space we can also count in optimal $O(m)$ time. No dictionary-compressed index had obtained this time before. All our indexes can be constructed in $O(n)$ space and $O(n\log n)$ expected time. As a byproduct of independent interest...

cs.DS↗

arXiv2019

On Longest Common Property Preserved Substring Queries

We revisit the problem of longest common property preserving substring queries introduced by~Ayad et al. (SPIRE 2018, arXiv 2018). We consider a generalized and unified on-line setting, where we are given a set $X$ of $k$ strings of total length $n$ that can be pre-processed so that, given a query string $y$ and a positive integer $k'\leq k$, we can determine the longest substring of $y$ that satisfies some specific property and is common to at least $k'$ strings in $X$. Ayad et al. considered the longest square-free substring in an on-line setting and the longest periodic and palindromic substring in an off-line setting. In this paper, we give efficient solutions in the on-line setting for finding the longest common square, periodic, palindromic, and Lyndon substrings. More precisely, we show that $X$ can be pre-processed in $O(n)$ time resulting in a data structure of $O(n)$ size that answers queries in $O(|y|\logσ)$ time and $O(1)$ working space, where $σ$ is the size of the alphabet, and the common substring must be a square, a periodic substring, a palindrome, or a Lyndon word.

cs.DS↗

arXiv2019

String Synchronizing Sets: Sublinear-Time BWT Construction and Optimal LCE Data Structure

Burrows-Wheeler transform (BWT) is an invertible text transformation that, given a text $T$ of length $n$, permutes its symbols according to the lexicographic order of suffixes of $T$. BWT is one of the most heavily studied algorithms in data compression with numerous applications in indexing, sequence analysis, and bioinformatics. Its construction is a bottleneck in many scenarios, and settling the complexity of this task is one of the most important unsolved problems in sequence analysis that has remained open for 25 years. Given a binary string of length $n$, occupying $O(n/\log n)$ machine words, the BWT construction algorithm due to Hon et al. (SIAM J. Comput., 2009) runs in $O(n)$ time and $O(n/\log n)$ space. Recent advancements (Belazzougui, STOC 2014, and Munro et al., SODA 2017) focus on removing the alphabet-size dependency in the time complexity, but they still require $Ω(n)$ time. In this paper, we propose the first algorithm that breaks the $O(n)$-time barrier for BWT construction. Given a binary string of length $n$, our procedure builds the Burrows-Wheeler transform in $O(n/\sqrt{\log n})$ time and $O(n/\log n)$ space. We complement this result with a conditional lower bound proving that any further progress in the time complexity of BWT construction would yield faster algorithms for the very well studied problem of counting inversions: it would improve the state-of-the-art $O(m\sqrt{\log m})$-time solution by Chan and Pǎtraşcu (SODA 2010). Our algorithm is based on a novel concept of string synchronizing sets, which is of independent interest. As one of the applications, we show that this technique lets us design a data structure of the optimal size $O(n/\log n)$ that answers Longest Common Extension queries (LCE queries) in $O(1)$ time and, furthermore, can be deterministically constructed in the optimal $O(n/\log n)$ time.

cs.DS↗