arXiv subjects

Philip Bille

Publications and source records attributed to Philip Bille.

At least 19 recordsLinked to original sources

arXiv2026

Average-Case Optimal Encodings and Efficient Worst-Case Indices for Element Distinctness Queries

We study the data structure version of the \emph{element distinctness problem}: preprocess an array of $n$ elements from an alphabet of size $\sigma$ to answer \textsc{All-Distinct} queries, asking whether a given range contains only distinct elements. We first focus on \emph{uniformly random arrays}: in the encoding model, where access to the input at query time is not allowed, we prove a lower bound on the expected space; for instance, the lower bound is $n$, $1.3627n$, $1.5153n$, $1.5824n$ bits for $\sigma = 2,3,4,5$, and approximately $n\sqrt{\pi/(2\sigma)}\,\log\sigma$ bits for $\sigma =\omega(1)$. We complement this by designing different average-case optimal encodings, supporting \textsc{All-Distinct} queries in worst-case time $O(1)$, $o(\log^{2}{\log{n}})$, or $O(\log\log{n})$ depending on $\sigma$, and $O(1)$ expected time for any $\sigma = \omega(1)$. We then switch to worst-case (non-random) arrays: in the indexing model, where access to the input is allowed, we prove a cell-probe space-time tradeoff lower bound showing that any index using $n/b$ bits must have $\Omega(b/\log{b})$ query time. We conclude by presenting a simple index almost matching this lower bound.

cs.DS↗

arXiv2026

Tighter Bounds for Wheeler Determinization

Given a Wheeler NFA $\mathcal{A}$, the Wheeler determinization problem is to construct a Wheeler DFA $\mathcal{D}$ that accepts the same language as $\mathcal{A}$. We use the notation $n_{\mathcal{A}},m_{\mathcal{A}}$ for the number of vertices and edges of $\mathcal{A}$, and equivalently $n_{\mathcal{D}},m_{\mathcal{D}}$ for $\mathcal{D}$. Alanko et al. [SODA 2020, Inf. Comp. 2021] solve this problem in $O(n_{\mathcal{A}}^3)$ time, by constructing a $\\mathcal{D}$ that always satisfies $n_\mathcal{D}\leq 2n_\mathcal{A} - 1$. In this paper, we show how to improve the running time to $O(n_{\mathcal{A}} + m_{\mathcal{A}} + n_{\mathcal{D}} + m_{\mathcal{D}})$ when the Wheeler order of $\mathcal{A}$ is given. If the Wheeler order is not present, we achieve $O(n_\mathcal{A} + m_\mathcal{A}\log n_\mathcal{A} + n_\mathcal{D} + m_\mathcal{D})$ time by using an algorithm of Becker et al. [ESA 2023]). Our running time is a factor $n_{\mathcal{A}}^2/\sigma$ faster than the state of the art for sorted inputs, where $\sigma$ is the size of the alphabet. Furthermore, for $\sigma=O(1)$ we have the first linear time algorithm for this problem. We show that our bound is tight with any combination of $n_\mathcal{A}$ and $\sigma$, by giving a family of inputs for which our output $\mathcal{D}$ is minimum, and of maximum size $\Theta(n_\mathcal{A}\sigma)$.

cs.DS↗

arXiv2026

Fast and Compact Graph Cuts for the Boykov-Kolmogorov Algorithm

Computing a minimum $s$-$t$ cut in a graph is a solution to a wide range of computer vision problems, and is often done using the Boykov-Kolmogorov (BK) algorithm. In this paper, we revisit the BK algorithm from both a theoretical and practical point of view. We improve the analysis of the time complexity of the BK algorithm to $O(mn|C|)$ and propose a new algorithm, the fast and compact BK (fcBK) algorithm, with a time complexity of $O(m|C|)$, where $m$, $n$, and $|C|$ are the number of edges, number of vertices, and the capacity of the cut, respectively. We additionally propose a compact graph representation that allows our implementation to find a minimum $s$-$t$ cut in a graph with upwards of $10^9$ vertices and $10^{10}$ edges on a machine with 128 GB of memory. We find our implementation of the BK algorithm to be the fastest available implementation of the BK algorithm when evaluating on a comprehensive set of benchmark datasets, highlighting the importance of memory-efficient implementations. We make our implementations publicly available for further research and implementation development within minimum $s$-$t$ cut algorithms.

cs.CV↗

arXiv2025

Improved Extended Regular Expression Matching

An extended regular expression $R$ specifies a set of strings formed by characters from an alphabet combined with concatenation, union, intersection, complement, and star operators. Given an extended regular expression $R$ and a string $Q$, the extended regular expression matching problem is to decide if $Q$ matches any of the strings specified by $R$. Extended regular expression matching was introduced by Hopcroft and Ullman in the 1970s, who gave a simple dynamic programming solution using $O(n^3m)$ time and $O(n^2m)$ space, where $n$ is the length of $Q$ and $m$ is the length of $R$. The current state-of-the art solution, by Yamamoto and Miyazaki uses $O(\frac{n^3k + n^2m}{w} + n + m)$ time and $O(\frac{n^2k + nm}{w} + n + m)$ space, where $k$ is the number of negation and complement operators in $R$ and $w$ is the number of bits in a machine word. This roughly replaces the $m$ factor with $k$ in the dominant terms of both the space and time bounds of the classical Hopcroft and Ullman algorithm. In this paper, we present a new solution that solves extended regular expression matching in \[ O\left(n^\omega k + \frac{n^2m}{\max(w/\log w, \log n)} + m\right) \] time and $O(\frac{n^2 \log k}{w} + n + m) = O(n^2 +m)$ space, where $\omega \approx 2.3716$ is the exponent of matrix multiplication. Essentially, this replaces the dominant $n^3k$ term with $n^\omega k$ in the time bound, while simultaneously improving the $n^2k$ term in the space to $O(n^2)$. Our results are based on a surprisingly simple combination of techniques and insights, including a compact representation to store and efficiently combine substring matches, a clustering technique for parse trees of extended regular expressions, and a new efficient combination of finite automaton simulation with our substring match representation to speed up the classic dynamic programming solution.

cs.DS↗

arXiv2025

Compressed Dictionary Matching on Run-Length Encoded Strings

Given a set of pattern strings $\mathcal{P}=\{P_1, P_2,\ldots P_k\}$ and a text string $S$, the classic dictionary matching problem is to report all occurrences of each pattern in $S$. We study the dictionary problem in the compressed setting, where the pattern strings and the text string are compressed using run-length encoding, and the goal is to solve the problem without decompression and achieve efficient time and space in the size of the compressed strings. Let $m$ and $n$ be the total length of the patterns $\mathcal{P}$ and the length of the text string $S$, respectively, and let $\overline{m}$ and $\overline{n}$ be the total number of runs in the run-length encoding of the patterns in $\mathcal{P}$ and $S$, respectively. Our main result is an algorithm that achieves $O( (\overline{m} + \overline{n})\log \log m + \mathrm{occ})$ expected time, and $O(\overline{m})$ space, where $\mathrm{occ}$ is the total number of occurrences of patterns in $S$. This is the first non-trivial solution to the problem. Since any solution must read the input, our time bound is optimal within an $\log \log m$ factor. We introduce several new techniques to achieve our bounds, including a new compressed representation of the classic Aho-Corasick automaton and a new efficient string index that supports fast queries in run-length encoded strings.

cs.DS↗

arXiv2024

Differentially Private Substring and Document Counting with Near-Optimal Error

For databases consisting of many text documents, one of the most fundamental data analysis tasks is counting (i) how often a pattern appears as a substring in the database (substring counting) and (ii) how many documents in the collection contain the pattern as a substring (document counting). If such a database contains sensitive data, it is crucial to protect the privacy of individuals in the database. Differential privacy is the gold standard for privacy in data analysis. It gives rigorous privacy guarantees, but comes at the cost of yielding less accurate results. In this paper, we carry out a theoretical study of substring and document counting under differential privacy. We propose a data structure storing $\epsilon$-differentially private counts for all possible query patterns with a maximum additive error of $O(\ell\cdot\mathrm{polylog}(n\ell|\Sigma|))$, where $\ell$ is the maximum length of a document in the database, $n$ is the number of documents, and $|\Sigma|$ is the size of the alphabet. We also improve the error bound for document counting with $(\epsilon, \delta)$-differential privacy to $O(\sqrt{\ell}\cdot\mathrm{polylog}(n\ell|\Sigma|))$. We show that our additive errors for substring counting and document counting are optimal up to an $O(\mathrm{polylog}(n\ell))$ factor both for $\epsilon$-differential privacy and $(\epsilon, \delta)$-differential privacy. Our data structures immediately lead to improved algorithms for related problems, such as privately mining frequent substrings and q-grams. Additionally, we develop a new technique of independent interest for differentially privately computing a general class of counting functions on trees.

cs.DS↗

arXiv2024

Succinct Data Structures for Segments

We consider succinct data structures for representing a set of $n$ horizontal line segments in the plane given in rank space to support \emph{segment access}, \emph{segment selection}, and \emph{segment rank} queries. A segment access query finds the segment $(x_1, x_2, y)$ given its $y$-coordinate ($y$-coordinates of the segments are distinct), a segment selection query finds the $j$th smallest segment (the segment with the $j$th smallest $y$-coordinate) among the segments crossing the vertical line for a given $x$-coordinate, and a segment rank query finds the number of segments crossing the vertical line through $x$-coordinate $i$ with $y$-coordinate at most $y$, for a given $x$ and $y$. This problem is a central component in compressed data structures for persistent strings supporting random access. Our main result is data structure using $2n\lg{n} + O(n\lg{n}/\lg{\lg{n}})$ bits of space and $O(\lg{n}/\lg{\lg{n}})$ query time for all operations. We show that this space bound is optimal up to lower-order terms. We will also show that the query time for segment rank is optimal. The query time for segment selection is also optimal by a previous bound. To obtain our results, we present a novel segment wavelet tree data structure of independent interest. This structure is inspired by and extends the classic wavelet tree for sequences. This leads to a simple, succinct solution with $O(\log n)$ query times. We then extend this solution to obtain optimal query time. Our space lower bound follows from a simple counting argument, and our lower bound for segment rank is obtained by a reduction from 2-dimensional counting.

cs.DS↗

arXiv2024

Dynamic Range Minimum Queries on the Ultra-Wide Word RAM

We consider the dynamic range minimum problem on the ultra-wide word RAM model of computation. This model extends the classic $w$-bit word RAM model with special ultrawords of length $w^2$ bits that support standard arithmetic and boolean operation and scattered memory access operations that can access $w$ (non-contiguous) locations in memory. The ultra-wide word RAM model captures (and idealizes) modern vector processor architectures. The goal in the dynamic range minimum problem is to maintain an array $A$ of $n$ $w$-bit integers subject to range minimum queries (given indices $i$ and $j$ return a smallest integer in the subarray $A[i..j]$) and updates (given index $i$ and integer $\alpha$ set $A[i] \leftarrow \alpha$). Our main result is a data structure that supports range minimum queries and updates in $O(\log \log \log n)$ time and uses $O(n/\log n)$ space in addition to the input array. This exponentially improves the time of existing techniques. Our result is based on a simple reduction to prefix minimum computations on sequences $O(\log n)$ words combined with a new parallel, recursive implementation of these.

cs.DS↗

arXiv2024

Toward Greener Matrix Operations by Lossless Compressed Formats

Sparse matrix-vector multiplication (SpMV) is a fundamental operation in machine learning, scientific computing, and graph algorithms. In this paper, we investigate the space, time, and energy efficiency of SpMV using various compressed formats for large sparse matrices, focusing specifically on Boolean matrices and real-valued vectors. Through extensive analysis and experiments conducted on server and edge devices, we found that different matrix compression formats offer distinct trade-offs among space usage, execution time, and energy consumption. Notably, by employing the appropriate compressed format, we can reduce energy consumption by an order of magnitude on both server and single-board computers. Furthermore, our experiments indicate that while data parallelism can enhance execution speed and energy efficiency, achieving simultaneous time and energy efficiency presents partially distinct challenges. Specifically, we show that for certain compression schemes, the optimal degree of parallelism for time does not align with that for energy, thereby challenging prevailing assumptions about a straightforward linear correlation between execution time and energy consumption. Our results have significant implications for software engineers in all domains where SpMV operations are prevalent. They also suggest that similar studies exploring the trade-offs between time, space, and energy for other compressed data structures can substantially contribute to designing more energy-efficient software components.

cs.DS↗

arXiv2024

Size-constrained Weighted Ancestors with Applications

The weighted ancestor problem on a rooted node-weighted tree $T$ is a generalization of the classic predecessor problem: construct a data structure for a set of integers that supports fast predecessor queries. Both problems are known to require $Ω(\log\log n)$ time for queries provided $\mathcal{O}(n\text{ poly} \log n)$ space is available, where $n$ is the input size. The weighted ancestor problem has attracted a lot of attention by the combinatorial pattern matching community due to its direct application to suffix trees. In this formulation of the problem, the nodes are weighted by string depth. This research has culminated in a data structure for weighted ancestors in suffix trees with $\mathcal{O}(1)$ query time and an $\mathcal{O}(n)$-time construction algorithm [Belazzougui et al., CPM 2021]. In this paper, we consider a different version of the weighted ancestor problem, where the nodes are weighted by any function $\textsf{weight}$ that maps the nodes of $T$ to positive integers, such that $\textsf{weight}(u)\le \textsf{size}(u)$ for any node $u$ and $\textsf{weight}(u_1)\le \textsf{weight}(u_2)$ if node $u_1$ is a descendant of node $u_2$, where $\textsf{size}(u)$ is the number of nodes in the subtree rooted at $u$. In the size-constrained weighted ancestor (SWA) problem, for any node $u$ of $T$ and any integer $k$, we are asked to return the lowest ancestor $w$ of $u$ with weight at least $k$. We show that for any rooted tree with $n$ nodes, we can locate node $w$ in $\mathcal{O}(1)$ time after $\mathcal{O}(n)$-time preprocessing. In particular, this implies a data structure for the SWA problem in suffix trees with $\mathcal{O}(1)$ query time and $\mathcal{O}(n)$-time preprocessing, when the nodes are weighted by $\textsf{weight}$. We also show several string-processing applications of this result.

cs.DS↗

arXiv2024

Gapped String Indexing in Subquadratic Space and Sublinear Query Time

In Gapped String Indexing, the goal is to compactly represent a string $S$ of length $n$ such that for any query consisting of two strings $P_1$ and $P_2$, called patterns, and an integer interval $[α, β]$, called gap range, we can quickly find occurrences of $P_1$ and $P_2$ in $S$ with distance in $[α, β]$. Gapped String Indexing is a central problem in computational biology and text mining and has thus received significant research interest, including parameterized and heuristic approaches. Despite this interest, the best-known time-space trade-offs for Gapped String Indexing are the straightforward $O(n)$ space and $O(n+occ)$ query time or $Ω(n^2)$ space and $\tilde{O}(|P_1| + |P_2| + occ)$ query time. We break through this barrier obtaining the first interesting trade-offs with polynomially subquadratic space and polynomially sublinear query time. In particular, we show that, for every $0\leq δ\leq 1$, there is a data structure for Gapped String Indexing with either $\tilde{O}(n^{2-δ/3})$ or $\tilde{O}(n^{3-2δ})$ space and $\tilde{O}(|P_1| + |P_2| + n^δ\cdot (occ+1))$ query time, where $occ$ is the number of reported occurrences. As a new tool towards obtaining our main result, we introduce the Shifted Set Intersection problem. We show that this problem is equivalent to the indexing variant of 3SUM (3SUM Indexing). Via a series of reductions, we obtain a solution to the Gapped String Indexing problem. Furthermore, we enhance our data structure for deciding Shifted Set Intersection, so that we can support the reporting variant of the problem. Via the obtained equivalence to 3SUM Indexing, we thus give new improved data structures for the reporting variant of 3SUM Indexing, and we show how this improves upon the state-of-the-art solution for Jumbled Indexing for any alphabet of constant size $σ>5$.

cs.DS↗

arXiv2024

The Fine-Grained Complexity of Episode Matching

Given two strings $S$ and $P$, the Episode Matching problem is to find the shortest substring of $S$ that contains $P$ as a subsequence. The best known upper bound for this problem is $\tilde O(nm)$ by Das et al. (1997) , where $n,m$ are the lengths of $S$ and $P$, respectively. Although the problem is well studied and has many applications in data mining, this bound has never been improved. In this paper we show why this is the case by proving that no $O((nm)^{1-ε})$ algorithm (even for binary strings) exists, unless the Strong Exponential Time Hypothesis (SETH) is false. We then consider the indexing version of the problem, where $S$ is preprocessed into a data structure for answering episode matching queries $P$. We show that for any $τ$, there is a data structure using $O(n+\left(\frac{n}τ\right)^k)$ space that answers episode matching queries for any $P$ of length $k$ in $O(k\cdot τ\cdot \log \log n )$ time. We complement this upper bound with an almost matching lower bound, showing that any data structure that answers episode matching queries for patterns of length $k$ in time $O(n^δ)$, must use $Ω(n^{k-kδ-o(1)})$ space, unless the Strong $k$-Set Disjointness Conjecture is false. Finally, for the special case of $k=2$, we present a faster construction of the data structure using fast min-plus multiplication of bounded integer matrices.

cs.DS↗

arXiv2024

String Indexing for Top-$k$ Close Consecutive Occurrences

The classic string indexing problem is to preprocess a string $S$ into a compact data structure that supports efficient subsequent pattern matching queries, that is, given a pattern string $P$, report all occurrences of $P$ within $S$. In this paper, we study a basic and natural extension of string indexing called the string indexing for top-$k$ close consecutive occurrences problem (SITCCO). Here, a consecutive occurrence is a pair $(i,j)$, $i < j$, such that $P$ occurs at positions $i$ and $j$ in $S$ and there is no occurrence of $P$ between $i$ and $j$, and their distance is defined as $j-i$. Given a pattern $P$ and a parameter $k$, the goal is to report the top-$k$ consecutive occurrences of $P$ in $S$ of minimal distance. The challenge is to compactly represent $S$ while supporting queries in time close to the length of $P$ and $k$. We give three time-space trade-offs for the problem. Let $n$ be the length of $S$, $m$ the length of $P$, and $ε\in(0,1]$. Our first result achieves $O(n\log n)$ space and optimal query time of $O(m+k)$. Our second and third results achieve linear space and query times either $O(m+k^{1+ε})$ or $O(m + k \log^{1+ε} n)$. Along the way, we develop several techniques of independent interest, including a new translation of the problem into a line segment intersection problem and a new recursive clustering technique for trees.

cs.DS↗

arXiv2024

String Indexing with Compressed Patterns

Given a string $S$ of length $n$, the classic string indexing problem is to preprocess $S$ into a compact data structure that supports efficient subsequent pattern queries. In this paper we consider the basic variant where the pattern is given in compressed form and the goal is to achieve query time that is fast in terms of the compressed size of the pattern. This captures the common client-server scenario, where a client submits a query and communicates it in compressed form to a server. Instead of the server decompressing the query before processing it, we consider how to efficiently process the compressed query directly. Our main result is a novel linear space data structure that achieves near-optimal query time for patterns compressed with the classic Lempel-Ziv compression scheme. Along the way we develop several data structural techniques of independent interest, including a novel data structure that compactly encodes all LZ77 compressed suffixes of a string in linear space and a general decomposition of tries that reduces the search time from logarithmic in the size of the trie to logarithmic in the length of the pattern.

cs.DS↗

arXiv2023

Rank and Select on Degenerate Strings

A 'degenerate string' is a sequence of subsets of some alphabet; it represents any string obtainable by selecting one character from each set from left to right. Recently, Alanko et al. generalized the rank-select problem to degenerate strings, where given a character $c$ and position $i$ the goal is to find either the $i$th set containing $c$ or the number of occurrences of $c$ in the first $i$ sets [SEA 2023]. The problem has applications to pangenomics; in another work by Alanko et al. they use it as the basis for a compact representation of 'de Bruijn Graphs' that supports fast membership queries. In this paper we revisit the rank-select problem on degenerate strings, introducing a new, natural parameter and reanalyzing existing reductions to rank-select on regular strings. Plugging in standard data structures, the time bounds for queries are improved exponentially while essentially matching, or improving, the space bounds. Furthermore, we provide a lower bound on space that shows that the reductions lead to succinct data structures in a wide range of cases. Finally, we provide implementations; our most compact structure matches the space of the most compact structure of Alanko et al. while answering queries twice as fast. We also provide an implementation using modern vector processing features; it uses less than one percent more space than the most compact structure of Alanko et al. while supporting queries four to seven times faster, and has competitive query time with all the remaining structures.

cs.DS↗

arXiv2023

Sparse Regular Expression Matching

A regular expression specifies a set of strings formed by single characters combined with concatenation, union, and Kleene star operators. Given a regular expression $R$ and a string $Q$, the regular expression matching problem is to decide if $Q$ matches any of the strings specified by $R$. Regular expressions are a fundamental concept in formal languages and regular expression matching is a basic primitive for searching and processing data. A standard textbook solution [Thompson, CACM 1968] constructs and simulates a nondeterministic finite automaton, leading to an $O(nm)$ time algorithm, where $n$ is the length of $Q$ and $m$ is the length of $R$. Despite considerable research efforts only polylogarithmic improvements of this bound are known. Recently, conditional lower bounds provided evidence for this lack of progress when Backurs and Indyk [FOCS 2016] proved that, assuming the strong exponential time hypothesis (SETH), regular expression matching cannot be solved in $O((nm)^{1-ε})$, for any constant $ε> 0$. Hence, the complexity of regular expression matching is essentially settled in terms of $n$ and $m$. In this paper, we take a new approach and introduce a \emph{density} parameter, $Δ$, that captures the amount of nondeterminism in the NFA simulation on $Q$. The density is at most $nm+1$ but can be significantly smaller. Our main result is a new algorithm that solves regular expression matching in $$O\left(Δ\log \log \frac{nm}Δ +n + m\right)$$ time. This essentially replaces $nm$ with $Δ$ in the complexity of regular expression matching. We complement our upper bound by a matching conditional lower bound that proves that we cannot solve regular expression matching in time $O(Δ^{1-ε})$ for any constant $ε> 0$ assuming SETH.

cs.DS↗

arXiv2023

Fast Practical Compression of Deterministic Finite Automata

We revisit the popular \emph{delayed deterministic finite automaton} (\ddfa{}) compression algorithm introduced by Kumar~et~al.~[SIGCOMM 2006] for compressing deterministic finite automata (DFAs) used in intrusion detection systems. This compression scheme exploits similarities in the outgoing sets of transitions among states to achieve strong compression while maintaining high throughput for matching. The \ddfa{} algorithm and later variants of it, unfortunately, require at least quadratic compression time since they compare all pairs of states to compute an optimal compression. This is too slow and, in some cases, even infeasible for collections of regular expression in modern intrusion detection systems that produce DFAs of millions of states. Our main result is a simple, general framework for constructing \ddfa{} based on locality-sensitive hashing that constructs an approximation of the optimal \ddfa{} in near-linear time. We apply our approach to the original \ddfa{} compression algorithm and two important variants, and we experimentally evaluate our algorithms on DFAs from widely used modern intrusion detection systems. Overall, our new algorithms compress up to an order of magnitude faster than existing solutions with either no or little loss of compression size. Consequently, our algorithms are significantly more scalable and can handle larger collections of regular expressions than previous solutions.

cs.DS↗

arXiv2023

Sliding Window String Indexing in Streams

Given a string $S$ over an alphabet $Σ$, the 'string indexing problem' is to preprocess $S$ to subsequently support efficient pattern matching queries, i.e., given a pattern string $P$ report all the occurrences of $P$ in $S$. In this paper we study the 'streaming sliding window string indexing problem'. Here the string $S$ arrives as a stream, one character at a time, and the goal is to maintain an index of the last $w$ characters, called the 'window', for a specified parameter $w$. At any point in time a pattern matching query for a pattern $P$ may arrive, also streamed one character at a time, and all occurrences of $P$ within the current window must be returned. The streaming sliding window string indexing problem naturally captures scenarios where we want to index the most recent data (i.e. the window) of a stream while supporting efficient pattern matching. Our main result is a simple $O(w)$ space data structure that uses $O(\log w)$ time with high probability to process each character from both the input string $S$ and the pattern string $P$. Reporting each occurrence from $P$ uses additional constant time per reported occurrence. Compared to previous work in similar scenarios this result is the first to achieve an efficient worst-case time per character from the input stream. We also consider a delayed variant of the problem, where a query may be answered at any point within the next $δ$ characters that arrive from either stream. We present an $O(w + δ)$ space data structure for this problem that improves the above time bounds to $O(\log(w/δ))$. In particular, for a delay of $δ= εw$ we obtain an $O(w)$ space data structure with constant time processing per character. The key idea to achieve our result is a novel and simple hierarchical structure of suffix trees of independent interest, inspired by the classic log-structured merge trees.

cs.DS↗