Sentence Boundaries Research Articles

VÍRGULAS EM TEXTOS DE ALUNOS DE 11 A 15 ANOS DO ENSINO FUNDAMENTAL II: UM ESTUDO LONGITUDINAL

ABSTRACT In this article, we investigate the presence and the absence of commas in a longitudinal sample of texts of the genre report written by 11 to 15 years-old students from a school in the countryside of São Paulo State. Our hypothesis is that the comma is used when syntactic boundaries match the prosodic ones. The prosodic constituents are formed from syntactic and phonological information, according to principles defined in the Prosodic Phonology theoretical approach. The characterization of the conventional commas and the non-conventional absences of commas was made from the identification of syntactic boundaries in which they should or could be placed, taking as reference a Brazilian Portuguese grammar. The results were that the conventional comma increases during the years of schooling, as expected. However, the presence and absence of comma rates changed depending on the type of syntactic structure and on the occurrence of intonational phrase boundaries. Comma occurred predominantly in positions where intonation phrase boundaries match major syntactic boundaries, as a sentence boundary. In turn, the comma systematically does not occur in positions where the syntactic and prosodic boundaries do not match.

Zone-Flooding as a Discursive Strategy of Czech Anti-System News Portals

Abstract: The anti-system media (ANTS)—known for spreading disinformation—might seem to "flood [the media] zone" with a chaotic multitude of information: truths, untruths, and half-truths alike. The main goal of this study is to find evidence of systematicity in this seeming chaos: persistent and recurring narrative lines that run through the media class irrespective of the news topic. Two empirical methods (Keyword Analysis and Market Basket Analysis) are applied to large data from Czech online media (all articles, regardless of topic, from 40 ANTS web portals over three months in 2020). ANTS' narratives are advanced by creating specific associations. The current approach is based on the idea that texts can be characterized with the help of conceptual associations, pursuing concepts which co-occur within the same text regardless of sentence or paragraph boundaries. This approach thus differs from the frequently-used strategy in discourse analysis of examining phenomena such as collocations, use of passive voice, or nominalization. The distinct properties of ANTS can be highlighted by contrasting it to the mainstream media class and to reader expectations in journalistic practice. The results, culled from servers including those not explicitly sponsored by the Kremlin, indicate that a schematic set of narrative lines permeate ANTS: a model of the world divided into the West (USA, NATO, and the EU) and Russia, in which the West has a negative image relative to that of Russia. These narrative lines lead to an argumentation for Czechia's separation from the West (Czexit, leaving NATO) and for alignment with Russia.

Sentence Boundary Disambiguation for Tibetan Based on Attention Mechanism at the Syllable Level

Tibetan is a low-resource language with few existing electronic reference materials. The goal of Tibetan sentence boundary disambiguation (SBD) is to segment long text into sentences, and it is the foundation for downstream tasks corpora building. This study implemented the Tibetan SBD at the syllable level to avoid word segmentation (WS) errors affecting the accuracy of SBD. Specifically, the attention mechanism is introduced based on a recurrent neural network (RNN) to study Tibetan SBD. The primary objective is to determine, using a trained model, whether the shad contained in Tibetan text is the ending of the sentence, and implement experiments on syllable embedding and component embedding to measure the model's performance. The highest accuracy for Tibetan syllable embedding and component embedding is 96.23% and 95.40 %, respectively, and the F1 score reaches 96.23% and 95.37%, respectively. The experimental results demonstrate that the proposed method can achieve better results than the established rule-based and statistical methods without considering various syntactic and part-of-speech (POS) tagging rules. German and English data from the Europarl corpus and Thai data from the IWSLT2015 corpus are validated to prove the models’ reliability and generalizability. The results demonstrate that this method is efficient not only for low-resource languages but also for high-resource languages. More importantly, we can formally apply the experimental results of this study to the research of downstream tasks, such as machine translation and automatic summarization.

Length, position, and functions of inter-clausal Chinese–English code-switching in a bilingual novel

AbstractBased on a written Chinese–English code-switching corpus compiled from a Singaporean bilingual novel, this paper reports on the distribution of instances of inter-clausal code-switching in terms of their length, the distribution of the position of instances of inter-clausal code-switching in mixed sentences, and the pragmatic functions of inter-clausal code-switching of different lengths in different sentence positions. The study yields four results. First, the distribution in terms of length fits Zipf’s Law well, and bilingual speakers attempt to avoid lengthy inter-clausal code-switching in their communication. Second, the majority of inter-clausal code-switching appears in sentence boundary positions. Third, the major pragmatic functions of shorter or longer instances of inter-clausal code-switching are turn-opening, turn-closing, explicitness, and accuracy. Finally, instances of inter-clausal code-switching with medium clause lengths have two main functions, which are the conveying of irony and emphasis.

Estimating Sentence-like Structure in Synthetic Languages Using Information Topology.

Estimating sentence-like units and sentence boundaries in human language is an important task in the context of natural language understanding. While this topic has been considered using a range of techniques, including rule-based approaches and supervised and unsupervised algorithms, a common aspect of these methods is that they inherently rely on a priori knowledge of human language in one form or another. Recently we have been exploring synthetic languages based on the concept of modeling behaviors using emergent languages. These synthetic languages are characterized by a small alphabet and limited vocabulary and grammatical structure. A particular challenge for synthetic languages is that there is generally no a priori language model available, which limits the use of many natural language processing methods. In this paper, we are interested in exploring how it may be possible to discover natural ‘chunks’ in synthetic language sequences in terms of sentence-like units. The problem is how to do this with no linguistic or semantic language model. Our approach is to consider the problem from the perspective of information theory. We extend the basis of information geometry and propose a new concept, which we term information topology, to model the incremental flow of information in natural sequences. We introduce an information topology view of the incremental information and incremental tangent angle of the Wasserstein-1 distance of the probabilistic symbolic language input. It is not suggested as a fully viable alternative for sentence boundary detection per se but provides a new conceptual method for estimating the structure and natural limits of information flow in language sequences but without any semantic knowledge. We consider relevant existing performance metrics such as the F-measure and indicate limitations, leading to the introduction of a new information-theoretic global performance based on modeled distributions. Although the methodology is not proposed for human language sentence detection, we provide some examples using human language corpora where potentially useful results are shown. The proposed model shows potential advantages for overcoming difficulties due to the disambiguation of complex language and potential improvements for human language methods.

Attentional Competition and Semantic Integration in Low- and High-Span Readers.

The goal of the current study is to investigate the effects of the distractive textual information on the activation of predictive inference online, and how the readers with high or low working memory capacity (WMC) differ in their online activation and text memory. To test the two hypothesis of attentional competition (AC) and semantic integration (SI), we conducted three experiments to investigate whether a local prediction (e.g., “The vase broke”) and a global prediction (e.g., “The wife left her husband”), both of which could be derived from the description of a critical event (e.g., “The angry husband throws the delicate porcelain vase against the brick wall”), are generated in the mind of the reader, and how this generation process is influenced by contextual and cognitive factors of the reader (e.g., working memory capacity). The results of Experiment 1 and 2 suggest that the elaboration of the global aspects in the narrative reduces the local prediction, but makes the global prediction more salient to occur. The evidence from Experiment 3 confirms the hypothesis that even automatic processes are constrained by distant contextual factors, in combination with differences in working memory, and examines how referentially local and global predictions are intertwined in text comprehension. Overall, these data support the immediate integration hypothesis across sentence boundaries at different representation levels (cf. Schmalhofer and Perfetti, 2007), as well as interaction assumptions of different processing levels within referentially local and referentially global processing contexts (cf. Yang et al., 2005).

Constructing theoretically informed measures of pause duration in experimentally manipulated writing

This paper argues that traditional threshold-based approaches to the analysis of pauses in writing fail to capture the complexity of the cognitive processes involved in text production. It proposes that, to capture these processes, pause analysis should focus on the transition times between linearly produced units of text. Following a review of some of the problematic features of traditional pause analysis, the paper is divided into two sections. These are designed to demonstrate: (i) how to isolate relevant transitions within a text and calculate their durations; and (ii) the use of mixture modelling to identify structure within the distributions of pauses at different locations. The paper uses a set of keystroke logs collected from 32 university students writing argumentative texts about current affairs topics to demonstrate these methods. In the first section, it defines how pauses are calculated using a reproducible framework, explains the distinction between linear and non-linear text transitions, and explains how relevant sections of text are identified. It provides Excel scripts for automatically identifying relevant pauses and calculating their duration. The second section applies mixture modelling to linear transitions at sentence, sub sentence, between-word and within-word boundaries for each participant. It concludes that these transitions cannot be characterised by a single distribution of “cognitive” pauses. It proposes, further, that transitions between words should be characterised by a three-component distribution reflecting lexical, supra-lexical and reflective processes, while transitions at other text locations can be modelled by two-component distributions distinguishing between fluent and less fluent or more reflective processing. The paper concludes by recommending that, rather than imposing fixed thresholds to distinguish processes, researchers should instead impose a common set of theoretically informed distributions on the data and estimate how the parameters of these distributions vary for different individuals and under different conditions.

Real-world sentence boundary detection using multitask learning: A case study on French

AbstractWe propose a novel approach for sentence boundary detection in text datasets in which boundaries are not evident (e.g., sentence fragments). Although detecting sentence boundaries without punctuation marks has rarely been explored in written text, current real-world textual data suffer from widespread lack of proper start/stop signaling. Herein, we annotate a dataset with linguistic information, such as parts of speech and named entity labels, to boost the sentence boundary detection task. Via experiments, we obtained F1 scores up to 98.07% using the proposed multitask neural model, including a score of 89.41% for sentences completely lacking punctuation marks. We also present an ablation study and provide a detailed analysis to demonstrate the effectiveness of the proposed multitask learning method.

A Rule-Based Grammar and Spell Checking

Rule-Based Grammar and spell checking is the task of checking the content of any specific language using a particular set of rules for their grammar and a set of spellings for spell checking. In this work, libraries and functions of Python useful for Natural Language Processing are used to perform grammar and Spell checking. Python Libraries NLTK (for Sentence Boundary Detection and POS tagging), pyEnchant (for Spell checking by comparing with a standard dictionary), and Inflect (for indefinite article). The article includes information about these libraries, techniques, and steps used for Grammar and Spell Check.

Predicting native Chinese readers’ perception of sentence boundaries in written Chinese texts

The notion of sentencehood in Mandarin Chinese is much less well-defined than in many other languages, with a block of clauses often joined by commas without conjunctions and with the period often occurring at the end of a block of clauses to indicate meaning completeness rather than the completeness of a sentential structure. The potential factors that may affect native Chinese speakers’ judgment of meaning completeness and perception of sentence boundaries have not yet been systematically examined. In light of this research gap, this study investigates the factors that may play a role in native Chinese speakers’ sentence boundary perception. To this end, we conducted text re-punctuation experiments in two separate groups, a training group and a testing group, using different stimuli texts. The stimuli texts were annotated with multiple levels of linguistic information to identify potentially relevant variables that could affect the participants’ sentence boundary perception. Logistic regression and the Bayesian statistical methods were applied to test the potential effects of multiple variables on the participants’ responses. The logistic regression model trained on the data from the training group achieved a high level of accuracy in predicting the responses by the testing group. The model revealed a more important role of semantic information than syntactic information in the participants’ sentence boundary perception. The implications of our findings for understanding the perception of Chinese sentence boundaries are discussed.

What makes the past perfect and the future progressive? Experiential coordinates for a learnable, context-based model of tense and aspect

AbstractWe examined how language supports the expression of temporality within sentence boundaries in English, which has a rich inventory of grammatical means to express temporality. Using a computational model that mimics how humans learn from exposure we explored what the use of different tense and aspect (TA) combinations reveals about the interaction between our experience of time and the cognitive demands that talking about time puts on the language user. Our model was trained onn-grams extracted from the BNC to select the TA combination that fits the context best. It revealed the existence of two different sub-systems within the set of TA combinations, a “simplex” one that is supported lexically and is easy to learn, and a “complex” one that is supported contextually and is hard to learn. The finding that some TA combinations are essentially lexical in nature necessitates a rethink of tense and aspect as grammatical categories that form the axes of the temporal system. We argue that the system of temporal reference may be more fruitfully thought of as the result of learning a system that is steeped in experience and organised along a number of functional principles.

Sentence Boundary Extraction from Scientific Literature of Electric Double Layer Capacitor Domain: Tools and Techniques

Given the growth of scientific literature on the web, particularly material science, acquiring data precisely from the literature has become more significant. Material information systems, or chemical information systems, play an essential role in discovering data, materials, or synthesis processes using the existing scientific literature. Processing and understanding the natural language of scientific literature is the backbone of these systems, which depend heavily on appropriate textual content. Appropriate textual content means a complete, meaningful sentence from a large chunk of textual content. The process of detecting the beginning and end of a sentence and extracting them as correct sentences is called sentence boundary extraction. The accurate extraction of sentence boundaries from PDF documents is essential for readability and natural language processing. Therefore, this study provides a comparative analysis of different tools for extracting PDF documents into text, which are available as Python libraries or packages and are widely used by the research community. The main objective is to find the most suitable technique among the available techniques that can correctly extract sentences from PDF files as text. The performance of the used techniques Pypdf2, Pdfminer.six, Pymupdf, Pdftotext, Tika, and Grobid is presented in terms of precision, recall, f-1 score, run time, and memory consumption. NLTK, Spacy, and Gensim Natural Language Processing (NLP) tools are used to identify sentence boundaries. Of all the techniques studied, the Grobid PDF extraction package using the NLP tool Spacy achieved the highest f-1 score of 93% and consumed the least amount of memory at 46.13 MegaBytes.

Second Language Learners’ Pauses Over Different Times Intervals in L2 Writing Essays: Evidence From a Keystroke Logging Program

Purpose. Several studies have been conducted to analyse students’ pauses during first language and/or second language writing to indicate the magnitude of the underlying cognitive processes learners have. Majority of studies have examined students’ pauses at a threshold 200ms. However, little is known about recording second language learners' pauses at different pauses’ times over different types of genres. The current investigation reports a case study of L2 learners’ cognitive processes by recording their pauses (<500ms, <1000ms, and <2000ms) during L2 writing in response to multiple genres prompts. 
 Design / methodology / approach. Twenty-five postgraduate students were asked to write three essays over three weeks, and their writing processes were recorded using a keystroke logging program (Inputlog, 7, Leijten & van Waes, 2013). Data was triangulated using a log file from the keystroke logging program, a process graph for writing behavior through different stages, and a visual video recording of their captured screens during writing behaviours. 
 Findings. Results found that the students paused over sentence and paragraph boundaries and their pauses between paragraphs were significantly higher in writing narrative essay than in their argumentative essays at pauses intervals <500 and <1000ms respectively, and in turn, their pauses between sentences in an argumentative essay were significantly higher than their pauses in a descriptive essay at <500, <1000 respectively. However, there were no significant differences across word boundaries over genre types. 
 Conclusions. The current study extends the previous literature in examining the underlying cognitive processes during L2 writing tasks as the trendy issue of psycholinguistics. Knowing the cognitive processes is crucial in diagnosing the students’ difficulties in writing L2 essays as advanced technology has the potential to explore intrusively the accurate cognitive processes learners involved during writing tasks.
 Originality / value. This paper is innovative in examining a state-of-the-art issue and has implications to the field of psycholinguistics.

A Context-Enhanced De-identification System.

Many modern entity recognition systems, including the current state-of-the-art de-identification systems, are based on bidirectional long short-term memory (biLSTM) units augmented by a conditional random field (CRF) sequence optimizer. These systems process the input sentence by sentence. This approach prevents the systems from capturing dependencies over sentence boundaries and makes accurate sentence boundary detection a prerequisite. Since sentence boundary detection can be problematic especially in clinical reports, where dependencies and co-references across sentence boundaries are abundant, these systems have clear limitations. In this study, we built a new system on the framework of one of the current state-of-the-art de-identification systems, NeuroNER, to overcome these limitations. This new system incorporates context embeddings through forward and backward n -grams without using sentence boundaries. Our context-enhanced de-identification (CEDI) system captures dependencies over sentence boundaries and bypasses the sentence boundary detection problem altogether. We enhanced this system with deep affix features and an attention mechanism to capture the pertinent parts of the input. The CEDI system outperforms NeuroNER on the 2006 i2b2 de-identification challenge dataset, the 2014 i2b2 shared task de-identification dataset, and the 2016 CEGS N-GRID de-identification dataset (p < 0.01). All datasets comprise narrative clinical reports in English but contain different note types varying from discharge summaries to psychiatric notes. Enhancing CEDI with deep affix features and the attention mechanism further increased performance.

Contexts and Consequences of Sentence Splitting in Translation (English-French-Czech)

The present paper examines the contexts and consequences of sentence splitting in English, Czech and French translated fiction. In the data extracted from a parallel (multilingual) corpus, we analyze first a language-specific context of sentence splitting (sententialization of non-finite verb forms in translations from English and French into Czech), and second, contexts of splitting occurring in all directions of translation. We conclude that sentence boundaries are usually introduced at the point of a sentence entailing the fewest modifications in the target sentence, especially between two coordinate clauses; and that a systematic sentence splitting, deeply modifying the style of the source text, involves the effect of simplification and normalization.

Indonesian Sentence Boundary Detection using Deep Learning Approaches

Detecting the sentence boundary is one of the crucial pre-processing steps in natural language processing. It can define the boundary of a sentence since the border between a sentence, and another sentence might be ambiguous. Because there are multiple separators and dynamic sentence patterns, using a full stop at the end of a sentence is sometimes inappropriate. This research uses a deep learning approach to split each sentence from an Indonesian news document. Hence, there is no need to define any handcrafted features or rules. In Part of Speech Tagging and Named Entity Recognition, we use sequence labeling to determine sentence boundaries. Two labels will be used, namely O as a non-boundary token and E as the last token marker in the sentence. To do this, we used the Bi-LSTM approach, which has been widely used in sequence labeling. We have proved that our approach works for Indonesian text using pre-trained embedding in Indonesian, as in previous studies. This study achieved an F1-Score value of 98.49 percent. When compared to previous studies, the achieved performance represents a significant increase in outcomes..

Translation correction of English phrases based on optimized GLR algorithm

Abstract Basic syntactic analysis refers to sentence-level syntactic analysis. In the process of developing the Mat Link English–Chinese machine translation system, the Generalized Maximum Likelihood Ratio algorithm was improved, and a basic English syntax analyzer for English–Chinese translation was designed and implemented. The analyzer approves the structure of the analysis table with a variety of export products, introduces the character mapping function to realize the automatic recognition of the sentence boundary, uses the children of the same level to describe the grammatical structure of the sentence, and realizes the proverb from the original sentence to the target sentence stage conversion. Finally, through the analysis of example sentences, the design concept and working process of the basic grammar are explained.

Semi-supervised Thai Sentence Segmentation Using Local and Distant Word Representations

A sentence is typically treated as the minimal syntactic unit used for extracting valuable information from a longer piece of text. However, in written Thai, there are no explicit sentence markers. We proposed a deep learning model for the task of sentence segmentation that includes three main contributions. First, we integrate n-gram embedding as a local representation to capture word groups near sentence boundaries. Second, to focus on the keywords of dependent clauses, we combine the model with a distant representation obtained from self-attention modules. Finally, due to the scarcity of labeled data, for which annotation is difficult and time-consuming, we also investigate and adapt Cross-View Training (CVT) as a semi-supervised learning technique, allowing us to utilize unlabeled data to improve the model representations. In the Thai sentence segmentation experiments, our model reduced the relative error by 7.4% and 10.5% compared with the baseline models on the Orchid and UGWC datasets, respectively. We also applied our model to the task of pronunciation recovery on the IWSLT English dataset. Our model outperformed the prior sequence tagging models, achieving a relative error reduction of 2.5%. Ablation studies revealed that utilizing n-gram presentations was the main contributing factor for Thai, while the semi-supervised training helped the most for English.

Syntactic and lexical features of adolescent L2 students’ academic writing

With a premium placed on academic writing in the U.S. secondary education context, adolescent L2 students are expected to develop advanced skills to analyze, interpret, and produce complex texts in a variety of content areas, which require proficiency in academic language. To enhance the academic literacy of developing L2 writers and to meet their linguistic needs, it is crucial to identify the unique language features of their academic writing. This study seeks to examine syntactic and lexical features of the text-based analytical essays written by Spanish-speaking L2 students (7th–12th grades) from a public school in a western state of the United States. Employing manual sentence coding and quantitative measures of selected linguistic variables from Coh- Metrix, we analyzed student texts (N = 86) to identify common linguistic patterns and to examine how these linguistic features predict writing quality. Our findings reveal that sentence boundary issues, lack of syntactic variety, the underuse of sophisticated subordination to show connections between ideas, and low use of advanced vocabulary are common in adolescent L2 students’ writing. The results of regression analyses show that syntactic complexity and lexical sophistication predict human-judged writing quality. Implications for pedagogy to address the linguistic needs of adolescent L2 students are discussed.

Sentence boundary detection of various forms of Tunisian Arabic

Sentence boundary detection (SBD) is an essential step for a very large number of natural language processing applications such as parsing, information retrieval, automatic summarization, machine translation, etc. In this paper, we tackle the problem of SBD of dialectal Arabic, especially for the Tunisian dialect. We compare the efficiency of three learning algorithms: Deep Neuronal Networks (DNN), Support Vector Machines (SVM) and Conditional Random Fields (CRF) to detect the boundaries of sentences written in different types of dialect. The best model achieved an F-measure of 84.37% using CRF which is a popular formalism for structured prediction in NLP and it has been widely applied in text segmentation.