Source String Research Articles

Dynamic code compression for JavaScript engine

AbstractWeb applications created using web languages such as HTML, CSS, and JavaScript are widely used regardless of execution environments, owing to their portability and popularity. Since JavaScript language is conventionally used for complex computations in large‐scale web apps, many optimization techniques have been proposed to accelerate the JavaScript performance. However, most of these optimizations speed up JavaScript engines at the expense of consuming more memory. Hence, memory consumption becomes an additional concern in the JavaScript field. Based on the research on memory status, we unearthed that a substantial portion of the heap memory is allocated for JavaScript source code, particularly in lightweight JavaScript engines, which may range from 13.2% to a maximal 52.5% of the entire heap. To resolve this memory issue, this article suggests a new memory optimization method, called dynamic code compression that dynamically compresses the source code and keeps its compressed form instead of the source string. A novel heuristic is proposed to timely compress the source code. We also re‐design the internal structure of the JavaScript engine to efficiently compress the source code without incurring any conflict. Using our code compression method, we could reduce the entire heap memory by up to 43.3% and consistently downsize the overall heap size. From an evaluation of standard benchmarks, our approach showed just 2.7% degradation in performance with negligible compression overhead, proving the feasibility of the code compression technique.

Features of gravitational interaction as the basis of “Dark Energy” in the gas of null strings

In this work, the influence of the shape of the null strings that form the gas on the gravitational properties and evolution of the null string gas was investigated. An interesting result of the study is that the test null string, being in the interaction zone, can only have a shape that exactly repeats the shape of the source string. As a result, null strings having different spatial forms can be located in the gas only outside the zone of interaction with each other. The “size” (width) of the zone of interaction of each null string in the gas is determined by the value of its initial momenta, and for some critical values it can occupy the entire space. The size (width) and “impenetrability” of the interaction zones for null strings having different spatial shapes can be a physical source of “strong” gravitational repulsion in the gas of null strings. In turn, the “strong” gravitational repulsion between null strings with different spatial shapes can be a source of a long-term accelerated expansion of such a gas (a source of “dark energy”).

Hierarchical Context Tagging for Utterance Rewriting

Utterance rewriting aims to recover coreferences and omitted information from the latest turn of a multi-turn dialogue. Recently, methods that tag rather than linearly generate sequences have proven stronger in both in- and out-of-domain rewriting settings. This is due to a tagger's smaller search space as it can only copy tokens from the dialogue context. However, these methods may suffer from low coverage when phrases that must be added to a source utterance cannot be covered by a single context span. This can occur in languages like English that introduce tokens such as prepositions into the rewrite for grammaticality. We propose a hierarchical context tagger (HCT) that mitigates this issue by predicting slotted rules (e.g., "besides _") whose slots are later filled with context spans. HCT (i) tags the source string with token-level edit actions and slotted rules and (ii) fills in the resulting rule slots with spans from the dialogue context. This rule tagging allows HCT to add out-of-context tokens and multiple spans at once; we further cluster the rules to truncate the long tail of the rule distribution. Experiments on several benchmarks show that HCT can outperform state-of-the-art rewriting systems by ~2 BLEU points.

On More General Distributions of Random Binning for Slepian–Wolf Encoding

Traditionally, ensembles of Slepian-Wolf (SW) codes are defined such that every bin of each $n$-vector of each source is randomly drawn under the uniform distribution across the sets $\{0,1,\ldots,2^{nR_X}-1\}$ and $\{0,1,\ldots,2^{nR_Y}-1\}$, where $R_X$ and $R_Y$ are the coding rates of the two sources, $X$ and $Y$, respectively. In a few more recent works, where only one source, say, $X$, is compressed and the other one, $Y$, serves as side information available at the decoder, the scope is extended to variable-rate S-W (VRSW) codes, where the rate is allowed to depend on the type class of the source string, but still, the random-binning distribution is assumed uniform within the corresponding, type-dependent, bin index set. In this expository work, we investigate the role of the uniformity of the random binning distribution from the perspective of the trade-off between the reliability (defined in terms of the error exponent) and the compression performance (measured from the viewpoint of the source coding exponent). To this end, we study a much wider class of random-binning distributions, which includes the ensemble of VRSW codes as a special case, but it also goes considerably beyond. We first show that, with the exception of some pathological cases, the smaller ensemble, of VRSW codes, is as good as the larger ensemble in terms the trade-off between the error exponent and the source coding exponent. Notwithstanding this finding, the wider class of ensembles proposed is motivated in two ways. The first is that it outperforms VRSW codes in the above-mentioned pathological cases, and the second is that it allows robustness: in the event of a system failure that causes unavailability of the compressed bit-stream from one of the sources, it still allows reconstruction of the other source within some controllable distortion.

A Current Allocation Strategy Based Balancing Technique of Voltage Source String in Switch-Ladder Inverter and Its Switched-Capacitor Variety

A charge balancing method for series voltage source feeding multilevel inverters (MLI) based on current allocation control is presented. Unlike those MLIs based on isolated dc sources, the series-connected elements in a voltage source string, such as a battery or supercapacitor stack, cannot be by-passed individually. This brings the problems of source imbalance and limited redundant switching states for current allocation control at high modulation index. In this article, the source balance in the switch-ladder inverter is achieved by allocating the current distribution. Furthermore, the balancing effectiveness is enhanced by a novel configuration combining a switched-capacitor (SC) circuit. The working principle and modulation index limit for the current allocation method is elucidated with a flow diagram and mathematical representation. The feasibility and operation of the proposed method and topology are successfully validated by a simulation study and the experimental results.

Lossless Event Compression of Discrete Event Systems

This article investigates the lossless event compression problem of discrete event systems which is, given a discrete event system and a source string generated by it, to find a minimal recoverable compressed string by removing as many events as possible. In order for the problem to be well post, two compression protocols are introduced. One requires that the last event is always kept. The other requires that, for any loop substring, at least one event is kept. We say a compressed string is recoverable if we can uniquely determine the source string based on the knowledge of the given discrete event system. We first construct an automaton to present all the possible source strings for a given compressed string. Based on the automaton, an algorithm is proposed to check whether the given compressed string is recoverable or not. We then propose an algorithm to calculate a minimal recoverable compressed string for a source string. The compressed string satisfies monotonicity which can significantly reduce the computational complexity of the algorithm. Finally, we use a practical example to illustrate these results.

Improved Switched-Capacitor Integrated Multilevel Inverter With a DC Source String

The integration of switched-capacitor techniques into multilevel inverters (MLIs) with a dc source string contributes to the boosting of voltage gain, renders it particularly attractive in reducing the number of input dc sources in the series string. However, the recent two-stage MLI topologies suffer from high voltage stress across the backend H-bridge. Addressing this concern, an improved topology, namely the one-switched-capacitor integrated MLI (1SCI-MLI) is proposed in this paper. An extended topology of 1SC-MLI is also established by appropriate incorporation of another switched-capacitor circuit. The two proposed topologies are endowed with voltage boosting capability. They also feature low switch count and low number of dc sources. More importantly, they resolve the high voltage stress problem in the existing counterparts. Their corresponding operational analysis and comparisons with recent MLI topologies are discussed. Simulation and experimental results from a laboratory prototype are presented to validate the effectiveness of the proposed topologies.

Optimal mean-based algorithms for trace reconstruction

In the (deletion-channel) trace reconstruction problem, there is an unknown $n$-bit source string $x$. An algorithm is given access to independent traces of $x$, where a trace is formed by deleting each bit of~$x$ independently with probability~$\delta$. The goal of the algorithm is to recover~$x$ exactly (with high probability), while minimizing samples (number of traces) and running time. Previously, the best known algorithm for the trace reconstruction problem was due to Holenstein~et~al.; it uses $\exp(\tilde{O}(n^{1/2}))$ samples and running time for any fixed $0 < \delta < 1$. It is also what we call a "mean-based algorithm", meaning that it only uses the empirical means of the individual bits of the traces. Holenstein~et~al.~also gave a lower bound, showing that any mean-based algorithm must use at least $n^{\tilde{\Omega}(\log n)}$ samples. In this paper we improve both of these results, obtaining matching upper and lower bounds for mean-based trace reconstruction. For any constant deletion rate $0 < \delta < 1$, we give a mean-based algorithm that uses $\exp(O(n^{1/3}))$ time and traces; we also prove that any mean-based algorithm must use at least $\exp(\Omega(n^{1/3}))$ traces. In fact, we obtain matching upper and lower bounds even for $\delta$ subconstant and $\rho := 1-\delta$ subconstant: when $(\log^3 n)/n \ll \delta \leq 1/2$ the bound is $\exp(-\Theta(\delta n)^{1/3})$, and when $1/\sqrt{n} \ll \rho \leq 1/2$ the bound is $\exp(-\Theta(n/\rho)^{1/3})$. Our proofs involve estimates for the maxima of Littlewood polynomials on complex disks. We show that these techniques can also be used to perform trace reconstruction with random insertions and bit-flips in addition to deletions. We also find a surprising result: for deletion probabilities $\delta > 1/2$, the presence of insertions can actually help with trace reconstruction.

Generalized Topology of a Hybrid Switched- Capacitor Multilevel Inverter for High- Frequency AC Power Distribution

The generalized topology of a hybrid switched-capacitor multilevel inverter (SCMLI) for high-frequency ac (HFAC) power distribution system is presented. By employing the series–parallel conversion technique, the proposed SCMLI can achieve self-balancing of capacitor voltages and voltage step-up with arbitrary numbers of SC cells. The output power quality is improved by the intermediate voltage levels provided by the bidirectional switched converter connected to the dc source string. Besides, several active switches can be replaced with diodes if the large phase difference between output current and voltage is not required. Since the voltage ripple magnitude of the SC units is inversely proportional to the fundamental output frequency, the proposed SCMLI is more preferable for HFac applications. This paper elucidates the working principle of the proposed inverter with the detailed circuit and switching states’ description under staircase modulation. The mathematical analyses associated with the capacitor voltage ripples, sizing, and power loss are also provided. A 19-level MLI with a voltage gain of three was realized by cooperating two SC units with a voltage source string consisting of three symmetric dc sources. The operation and performance of the proposed inverter have been validated by the simulation and experiment on the inverter. The results showed that the inverter performed better at HFac condition and the power quality was satisfactory with the 19-level staircase output modulated at the fundamental frequency.

Dynamic Relative Compression, Dynamic Partial Sums, and Substring Concatenation

Given a static reference string R and a source string S, a relative compression of S with respect to R is an encoding of S as a sequence of references to substrings of R. Relative compression schemes are a classic model of compression and have recently proved very successful for compressing highly-repetitive massive data sets such as genomes and web-data. We initiate the study of relative compression in a dynamic setting where the compressed source string S is subject to edit operations. The goal is to maintain the compressed representation compactly, while supporting edits and allowing efficient random access to the (uncompressed) source string. We present new data structures that achieve optimal time for updates and queries while using space linear in the size of the optimal relative compression, for nearly all combinations of parameters. We also present solutions for restricted and extended sets of updates. To achieve these results, we revisit the dynamic partial sums problem and the substring concatenation problem. We present new optimal or near optimal bounds for these problems. Plugging in our new results we also immediately obtain new bounds for the string indexing for patterns with wildcards problem and the dynamic text and static pattern matching problem.

The Risks In Translating Smartphone Strings For Medical Study Smartphones Apps &amp; Smartwatches

Recently there has been an increase in the ‘Bring Your Own Device’ approach to clinical studies as a means of participants carrying out Patient Reported Outcomes (PROs) on their own smartphones or smartwatches. The aim of this study is to evaluate the risks in translating smartphone strings used in such apps and to highlight solutions in conveying the context of individual strings. We will focus on a study related to mobility being translated into France-French. This Linguistic Validation project underwent the following steps outlined by the ISPOR Principles of Good Practice: Preparation (including concept elaboration); Dual Forward Translation; Reconciliation; Editing; Dual Back Translation; Back Translation Review and Proofreading. During Back Translation Review, the risks in translating smartphone strings were analysed, establishing which translations needed further clarification and updates. When reviewing the Back Translations, 7% of items required updates. Of these; 82% were misinterpretations of the source due to lack of context and 18% were formatting updates to ensure the translations were short enough to fit within a smartphone screen. Without being able to refer to the English source in screenshot form, translating these smartphone strings into French presented two specific areas of risks. 1) Many translations were too long due to additional clarifications being required in French. 2) French occasionally required several translations depending on the specific context of each source string (e.g. ‘skip’ was both ‘ignoré’ and ‘non prise’). To provide clear and accurate text for the app, it is important to provide linguists with screenshots and concept elaborations from project initiation. This aids the linguist to retain the appropriate character length for the smart-device displays. French texts often require elaboration so developers should consider flexibility and precision of language when developing the app. Both elements ensure accurate Linguistic Validation of smartphone strings and a reduced need for edits.

Toward a theory of input-driven locally parsable languages

If a context-free language enjoys the local parsability property then, no matter how the source string is segmented, each segment can be parsed independently, and an efficient parallel parsing algorithm becomes possible. The new class of locally chain parsable languages (LCPLs), included in the deterministic context-free language family, is here defined by means of the chain-driven automaton and characterized by decidable properties of grammar derivations. Such automaton decides whether to reduce or not a substring in a way purely driven by the terminal characters, thus extending the well-known concept of input-driven (ID) alias visibly pushdown machines. The LCPL family extends and improves the practically relevant Floyd's operator-precedence (OP) languages which are known to strictly include the ID languages, and for which a parallel-parser generator exists.

On the Parameterised Complexity of String Morphism Problems

Given a source string u and a target string w, to decide whether w can be obtained by applying a string morphism on u (i. e., uniformly replacing the symbols in u by strings) constitutes an ����$\mathcal {NP}$-complete problem. We present a multivariate analysis of this problem (and its many variants) from the viewpoint of parameterised complexity theory, thereby pinning down the sources of its computational hardness. Our results show that most parameterised variants of the string morphism problem are fixed-parameter intractable and, apart from some very special cases, tractable variants can only be obtained by considering a large part of the input as parameters, namely the length of w and the number of different symbols in u.

Using swaps and deletes to make strings match

Given a source and a target string, their swap–delete–edit distance is the minimum number of interchange-consecutive-symbols and delete operations applied to the source string to make it equal the target string. We show that the swap–delete–edit distance of strings over alphabets of bounded size can be computed in polynomial time.

Nonblock Coding of Binary Sources by Probabilistic Enumeration

A simple and efficient nonblock coding scheme for binary sources is suggested. It uses a probability model to map any source string to a unique integer, and thereby defines an enumeration of all possible strings. Contrary to normal enumerative coding, the new method is noncombinatorial and operates sequentially, incrementing the code value symbol by symbol, by simple arithmetic. It is akin to arithmetic coding but does not use intervals. For memoryless sources, the redundancy per symbol is shown to be asymptotically less than p3 bits, where p is the smaller of symbol probabilities. Especially, for integer-valued 1/p , the code is asymptotically optimal. These results are confirmed by both analysis and experiments with arbitrary-precision arithmetic. In a practical implementation, the source is partitioned into substrings enabling restricted-precision arithmetic. Thanks to subtle implicit coding, the additional redundancy is marginal. The source model can be also context-based, and even adaptivity can be incorporated. A peculiarity of the method is that decoding is done backwards (LIFO). Hence, for higher-order models, the string suffix, not prefix, is used as the context domain. The speed of the practical version is close to that of binary arithmetic coders.

Tunstall Code, Khodak Variations, and Random Walks

A variable-to-fixed length encoder partitions the source string into variable-length phrases that belong to a given and fixed dictionary. Tunstall, and independently Khodak, designed variable-to-fixed length codes for memoryless sources that are optimal under certain constraints. In this paper, we study the Tunstall and Khodak codes using variety of techniques ranging from stopping times for sums of independent random variables to Tauberian theorems and Mellin transform. After proposing an algebraic characterization of the Tunstall and Khodak codes, we present new results on the variance and a central limit theorem for dictionary phrase lengths. This analysis also provides a new argument for obtaining asymptotic results about the mean dictionary phrase length and average redundancy rates.

Efficient algorithms for analyzing segmental duplications with deletions and inversions in genomes

BackgroundSegmental duplications, or low-copy repeats, are common in mammalian genomes. In the human genome, most segmental duplications are mosaics comprised of multiple duplicated fragments. This complex genomic organization complicates analysis of the evolutionary history of these sequences. One model proposed to explain this mosaic patterns is a model of repeated aggregation and subsequent duplication of genomic sequences.ResultsWe describe a polynomial-time exact algorithm to compute duplication distance, a genomic distance defined as the most parsimonious way to build a target string by repeatedly copying substrings of a fixed source string. This distance models the process of repeated aggregation and duplication. We also describe extensions of this distance to include certain types of substring deletions and inversions. Finally, we provide a description of a sequence of duplication events as a context-free grammar (CFG).ConclusionThese new genomic distances will permit more biologically realistic analyses of segmental duplications in genomes.

Efficient dictionary-based text rewriting using subsequential transducers

AbstractProblems in the area of text and document processing can often be described astext rewriting tasks: given an input text, produce a new text by applying some fixed set of rewriting rules. In its simplest form, a rewriting rule is given by a pair of strings, representing a source string (the “original”) and its substitute. By a rewriting dictionary, we mean a finite list of such pairs; dictionary-based text rewriting means to replace in an input text occurrences of originals by their substitutes. We present an efficient method for constructing, given a rewriting dictionaryD, a subsequential transducerthat accepts any texttas input and outputs the intended rewriting result under the so-called “leftmost-longest match” replacement with skips,t'. The time needed to compute the transducer is linear in the size of the input dictionary. Given the transducer, any texttof length |t| is rewritten in a deterministic manner in timeO(|t|+|t'|), wheret' denotes the resulting output text. Hence the resulting rewriting mechanism is very efficient. As a second advantage, using standard tools, the transducer can be directly composed with other transducers to efficiently solve more complex rewriting tasks in a single processing step.

Communicating Probability Distributions

A rate distortion problem is solved that is motivated by a quantum data compression problem. The goal is to send information about a source string x so that a receiver can construct a second string y for which the joint empirical probability distribution of x and y is close to some desired distribution. The problem differs from the usual rate distortion problems in that one must consider both remote sources and distortion functions that are not averages of per-letter distortion functions

자율 학습을 이용한 선형 정렬 말뭉치 구축

본 논문에서는 자을 선형 정렬 알고리즘을 이용하여 선형 정렬 말뭉치를 구축하는 방법을 제안한다. 기존의 자율 선형 정렬 알고리즘을 이용하여 선형 정렬 말뭉치를 구축할 경우, 두 문자열의 길이가 서로 다르면 정렬된 두 문자열(입력열과 출력열)에 모두 공백문자가 나타난다. 이 방법을 그대로 사용하면 정렬 말뭉치의 구축은 용이하나 정렬된 말뭉치를 이용하는 응용 시스템에서는 탐색 공간이 기하급수적으로 늘어날 뿐 아니라 구축된 정렬 말뭉치는 다양한 기계학습 방법에 두루 사용될 수 없다는 문제가 있다. 본 논문에서는 이들 문제를 최소화하기 위해서 입력열에는 공백문자가 나타나지 않도록 기존의 자을 선형 정렬 알고리즘을 수정하였다. 이 알고리즘을 이용해서 한영 음차 표기 및 복원, 영어 단어의 발음 생성, 영어 발음의 단어 생성, 한국어 형태소 분리 및 복원을 위한 정렬 말뭉치를 구축하였으며, 간단한 실험을 통해, 그들의 실용성을 입증해 보였다. In this paper, we propose a modified unsupervised linear alignment algorithm for building an aligned corpus. The original algorithm inserts null characters into both of two aligned strings (source string and target string), because the two strings are different from each other in length. This can cause some difficulties like the search space explosion for applications using the aligned corpus with null characters and no possibility of applying to several machine learning algorithms. To alleviate these difficulties, we modify the algorithm not to contain null characters in the aligned source strings. We have shown the usability of our approach by applying it to different areas such as Korean-English back-trans literation, English grapheme-phoneme conversion, and Korean morphological analysis.