Text Complexity Research Articles

A Method for Measuring Word Sequence Complexity of Text

ABSTRACT In linguistics and natural language processing, measuring text features is crucial for representing and revealing the properties of texts in terms of topics, genres, sentiment and more. Current methods predominantly rely on the frequency of linguistic units and rarely consider the syntagmatic properties of these units, which can reflect deeper linguistic characteristics (e.g. syntax and pragmatics). This paper proposes a general method for calculating the sequence complexity of text. By combining the features of word length and word frequency, two specific implementations of this method are provided. Using these two formulas, word sequence complexity of text based on the Brown Corpus and the Gutenberg Corpus is calculated. The results show that word sequence complexity of text indicates a characteristic of gradual stabilization. By comparing random texts with natural texts, it is found that word sequence complexity is influenced by syntactic rules rather than sentence order, and natural texts tend to alternate between words of varying lengths and frequencies. Classification experiment results indicate that the proposed word sequence complexity outperforms commonly used quantitative indicators for representing the genre of text, such as TTR, information entropy, Zipf’s law parameters and motifs.

Complexity and accuracy of verbal morphology in written L2 Italian

Abstract The present study aims to verify previous findings on the role of proficiency in the degree of complexity and accuracy of verbal morphemes in written essays from intermediate and advanced L2 Italian learners. In addition, by taking the perspective of usage-based theories on the distributional properties in the input, it investigates the extent to which contingency affects morphological complexity and accuracy by the emergence of cue-outcome associations. Hypotheses on the effects of contingency and proficiency on complexity and accuracy, and on their mutual relationships, are verified by adopting a confirmatory approach and by using Structural equation modeling. Results support the claims for non-universal mechanisms in the acquisition of verbal morphology: proficiency does not affect the morphological complexity of learner texts in the same way in inflectionally rich and poor languages, and the cue-outcome mechanisms of associative learning work to different extents depending on the inflectional systems of the target languages.

Syntax, stress and cognitive load, or on syntactic processing in simultaneous interpreting

Abstract This corpus-based study examines the effect of syntactic complexity in the source language on simultaneous interpreters’ cognitive load and stress. Previous studies show contrasting results regarding the source text syntax and cognitive load in interpreting, while the link between syntactic complexity and interpreters’ stress remains unexplored. Our research aims to fill this gap by measuring cognitive load through filled and silent pauses, and stress through fundamental frequency. We evaluated syntactic complexity by calculating dependency distance, i.e., the number of words between syntactically dependent elements. We used PINC, the Polish Interpreting Corpus (Chmiel et al. 2022), as our dataset. We found that higher syntactic complexity in the source text increases stress and cognitive load among interpreters. Additionally, when interpreters produce complex syntax themselves, their stress increases. The study provides empirical evidence for simplification in interpreting and shows that syntactic complexity in interpreters’ own output does not necessarily lead to increased cognitive load.

Online Plain Language Tool and Health Information Quality

Complex and ineffective health communication is a critical source of health inequity and occurs despite repeated policy directives to provide health information that is easy to understand and applies health literacy principles. To evaluate the effectiveness of the Sydney Health Literacy Lab Health Literacy Editor, an easy-to-use online plain language tool that supports health information providers to apply health literacy guidelines to written health information. This randomized clinical trial, conducted online in Australia from May 2023 to February 2024, included a convenience sample of health information providers with no previous experience using the Health Literacy Editor. Analysts were blinded to study group. Participants were randomized 1:1 to the intervention or control group. Participants in the intervention group were provided access to the Health Literacy Editor and a 30-minute online training program prior to editing 3 prespecified health texts. The Health Literacy Editor gives objective, real-time feedback on words and sentences. Control participants revised the texts using their own standard health information development processes. The preregistered primary outcome was the text school grade reading score (using a validated instrument, the Simple Measure of Gobbledygook). Secondary outcomes were text complexity (percentage of text using complex language), use of passive voice (number of instances), and subjective expert ratings (5-point Likert scale corresponding to items on the Patient Education Materials Assessment Tool). A total of 211 participants were randomized, with 105 in the intervention group and 106 in the control group. Of 181 participants in the intention-to-treat analysis (mean [SD] age, 41.0 [11.6] years; 154 women [85.1%]), 86 were in the intervention group and 95 in the control group. Texts revised in the intervention group had significantly improved grade reading scores (mean difference [MD], 2.48 grades; 95% CI, 1.84-3.12 grades; P < .001; Cohen d, 0.99), lower text complexity scores (MD, 6.86; 95% CI, 4.99-8.74; P < .001; Cohen d, 0.95), and less use of passive voice (MD, 0.95 instances; 95% CI, 0.44-1.47 instances; P < .001; Cohen d, 0.53) compared with texts revised in the control group in intention-to-treat analyses. Experts rated texts in the intervention group more favorably for word choice and style than those in the control group (MD, 0.44; 95% CI, 0.25-0.63; P < .001; Cohen d, 0.63), with no loss of meaning or content. In this randomized clinical trial, the Health Literacy Editor helped users simplify health information and apply health literacy guidelines to written text. The findings suggest the tool has high potential to improve development of health information for people who have low health literacy. As an online tool, the Health Literacy Editor is also easy to access and implement at scale. ANZCTR Identifier: ACTRN12623000386639.

Filologie – věc veřejná

Alexandr Stich theoretically contributed to the definition of the journalistic style in the system of functional styles, understanding it as a style primarily determined by the persuasive function. After 1989, he himself became a prominent publicist who tried to influence public opinion on matters of cultural and political importance in his texts published especially in Literární noviny, Lidové noviny and Přítomnost. In our analysis, we approach Stich’s journalism from the perspective of Hausenblas’s functional stylistics as a text complex, i.e. as a whole, which – as we show – is unified in function, style and meaning and is internally interconnected on the linguistic, thematic and compositional levels.

Об изменении подходов к формированию читательской грамотности школьников

The article highlights the results of a study of the problem of developing reading literacy among schoolchildren in Kyrgyzstan in the light of the requirements of the All-Republican Testing (ORT). ORT tasks place higher demands than subject standards in terms of students' language skills and level of text complexity, as well as in terms of depth of analysis, analytical, critical thinking and the ability to apply knowledge in practical scenarios mastered in school subject lessons. However, there are gaps between the requirements of ORT and teaching students reading and comprehension at school. To eliminate the gaps between subject standards, methods of teaching reading and comprehension at school and the requirements of ORT, the author proposes ways to solve the problem that will contribute to the development of students’ ability to read and understand a variety of texts of varying complexity, which, in turn, will contribute to the formation of students’ information skills. competence.

Uncovering condition information loss in medical text extraction: The challenge of non-contiguous spans

We investigated the limitations of conventional named entity recognition (NER) and entity linking (EL) methods in accurately extracting patient condition information from medical texts, focusing on the challenges posed by non-contiguous spans and the potential information loss. We utilized a corpus with entity-relation annotations, analyzing the frequency and nature of non-contiguous spans that include irrelevant entities within gaps. The corpus was further analyzed to pinpoint the types of entity representations predominantly linked with peripheral spans—those not encompassing central symptom-describing terms—with a focus on items, body parts, and clinical tests. Our analysis revealed that 18.6 % of patient condition expressions were non-contiguous spans containing irrelevant entities, suggesting an accuracy ceiling of 81.4 % for conventional NER and EL approaches in the worst-case scenario. The study highlights the importance of entity types such as items, body parts, and clinical tests in these expressions, indicating that conventional extraction methods incur considerable information loss. The findings underscore the need for more sophisticated information extraction techniques capable of handling the complexities of medical texts, including non-contiguous spans. Adapting methods that allow gaps within entities or employing graph-based term assignments can enhance the accuracy and comprehensiveness of medical text annotation.

Text complexity increase in Russian texts as a function of morphological changes

Text complexity increase in Russian texts as a function of morphological changes

Simulating judicial trial logic: Dual residual cross-attention learning for predicting legal judgment in long documents

Legal Judgment Prediction (LJP) plays a vital role in judicial assistance systems, aiming to predict judgment outcomes automatically from the fact descriptions in legal cases. A key challenge in LJP lies in effectively capturing decisive information that influences legal judgments — such as criminal events, behaviors, and consequences — especially from lengthy legal documents. Existing approaches, which incorporate domain-specific knowledge, have attempted to improve the ability to capture decisive information but often fail to adequately address the complexities of longer texts and may introduce noise, leading to incorrect predictions. To overcome these limitations, we propose a novel method, JuriSim, for predicting Chinese criminal legal judgments in long documents by incorporating the knowledge of judicial trial logic. Specifically, JuriSim extracts legal events and generates rationales based on fact descriptions to capture the decisive information that influences judgment outcomes. Then, a dual residual cross-attention mechanism is introduced to interactively process facts, events, and rationales for predicting relevant legal statutes, charges, and term of penalties. This mechanism allows the model to reduce the loss of important information and the retention of incorrect information during the aggregation process. Furthermore, we present a constrained cross-entropy loss, utilizing the topological relationship between charges, terms, and applicable law statutes. Experiments conducted on the publicly available CAIL-Long criminal dataset demonstrate the efficiency of the JuriSim framework in predicting legal judgments, especially for cases involving long documents. JuriSim_Lawformer shows a relative improvement of 3.45% in Macro-F1 for charge prediction and 3.05% for term of penalty prediction, compared to Lawformer.

Диагностика сформированности смыслового чтения при работе учеников средней школы с учебными математическими текстами

Learning is working with a variety of educational texts based on reading, understanding, evaluating and interpreting information. To identify the result of reading an educational mathematical text and the quality of understanding the meaning of the text space created by educational materials, it is necessary to diagnose reading, namely, the level of semantic reading. In the modern study of reading, the PISA international monitoring, the concept of reading literacy, its components and levels are defined. The success of understanding more complex educational mathematical texts will depend on the level of reader literacy. If the levels of reading literacy have evaluation criteria and diagnostic apparatus, then how to assess the level of semantic reading? The article examines the relationship between the levels of reading literacy and text comprehension, models for assessing the semantic complexity of the text and semantic reading skills, various level structures for diagnosing the understanding of a mathematical educational text. Based on the studied scientific material, the complexity of the educational mathematical text &amp;quot;Triangle&amp;quot; was assessed according to criteria such as the number of words, simple words, scientific terms and their complexity, the number of mentions of scientific terms in each paragraph. Diagnostic material has also been created to check the understanding of this text, each task of which is aimed at determining the understanding of each information block and assessing the degree of development of semantic reading.

Analysis of Textual Complexity in Danish News Articles on Climate Change

Structural linguistic features are often overlooked yet potentially important aspects of journalistic practice. Especially in news reporting on climate change, these features can play a crucial role as the proper use of language is tied to message credibility, processing fluency and knowledge retention, which can positively influence the reader to take more climate action. This article analyzes language use in Danish news articles on climate change using a sample of around 32,000 articles from four different outlet types (quality news, niche papers, tabloids, and public service broadcasters) published from 1990 to 2021. We create a machine-learning model of text complexity covering this concept’s semantic and syntactic dimensions. Our findings confirm expected differences in complexity between news outlets, highlighting tabloid articles as engaging with higher semantic complexity, while quality papers and niche papers exhibit higher syntactic complexity. We observe a significant decrease in semantic complexity and a slight increase in syntactic complexity over time, a trend towards more generic language, and an increased use of pronouns, verbs, and adverbs. Most of these changes can be attributed to the emergence of articles by public service broadcasters. Articles by public service broadcasters are characterised by high syntactic complexity, which we consider problematic due to their popularity among the general public.

“Making Sure the Books Check Every Box”: Preservice Teachers Learning to Select Texts for Reading Instruction

ABSTRACT A key consideration in providing effective reading instruction is the selection of a text. However, little is known about how preservice teachers are prepared to understand text complexity and how they leverage that knowledge in their text selection practices. Accordingly, this design/development study explored how preservice teachers (n = 81) gained declarative knowledge of text complexity, procedural knowledge of text selection practices, and conditional knowledge of activity, reader, and text considerations in their teacher education program. Findings indicate that following a two-course reading methods sequence in which they selected texts for small group instruction, typically performing preservice teachers gave the most attention to activity considerations (e.g. lesson-to-text match, instructional sequence) and the least attention to reader considerations (e.g. student interest). When noting text features teachers considered important, they described word-level features such as multisyllabic words, vocabulary, and decodability as most influencing text complexity. Furthermore, preservice teachers who held more declarative knowledge of text complexity tended to be more critical of text leveling systems whereas those with less knowledge of text complexity tended to view levels as an acceptable method of selecting texts. Together these findings suggest an emergent learning continuum on the topic of text selection for teacher education.

EVALUATING INDONESIAN HERO STORIES FEATURED IN CHILDREN’S LITERATURE

This study delves into elementary school teachers’ evaluation of children’s books featuring Indonesian national heroes. Despite the significant roles of historical figures picturebooks, research on how teachers incorporate it as useful resources in teaching is limited. In fact, the availability of quality historical fiction picturebooks that portray stories of national heroes in Indonesia is also scarce. Through snowballing technique, 25 teachers voluntarily participated. Through an online questionnaire, they were asked to provide their opinion about three Indonesian hero storybooks, which include (1) Aku Kartini [I am Kartini], (2) Wahidin Soedirohoesodo Sang Dokter Bangsa [Wahidin Soedirohoesodo the Doctor of The Nation], and (3) Menyusuri Tempat Tinggal Srikandi Aceh Cut Nyak Meutia [Strolling the Residence of an Acehnese Heroine Cut Nyak Meutia]. The participants were asked to examine aspects like text complexity, plot, illustration, moral values, and pedagogical strategies for classroom implementation. The data were analyzed qualitatively to describe the teachers’ evaluation on the quality of the books. The findings emphasize the significance of continuous teachers’ professional development, especially in book selection and utilization. The consensus surrounding Book #1 “Aku Kartini” suggests its potential as a valuable resource for teaching national heroes, as well as a model of quality historical fiction picturebook. We argue that adult’s evaluation, particularly teachers, in selecting books for children to read, play a pivotal role both in the children and in the practice of teaching.

In the Belly of Text Complexity: Unravelling the Nexus between Lexical Density and Readability

This study set out to explore the link between lexical density (LD) as a tool for determining text complexity and the general readability of texts (in this study, texts assigned to Senior High School learners in Ghana). To clearly establish this link, the levels of LD of assigned texts have to be first determined, and then, compared with their corresponding readability values. To that end, Ure’s (1971) LD formula and Halliday’s (1985b) LD formula have been used to determine the LD values of the sampled texts. Gunning’s (1952) and Flesch’s (1948) readability indices have been used to determine the readability values of same assigned texts. LD values for the sampled texts were found to be generally very high (above the widely agreed minimum of 40%); as their corresponding readability values were low. The correlation between LD and readability is therefore adequately revealed: high LD translates directly into low readability while low LD translates directly into high readability. Implicatively, low readability of texts definitely affects learner progress negatively. Keywords: lexical density, readability, Senior High School learners in Ghana, Ure’s (1971) LD formula, Halliday’s (1985) LD formula, Gunning’s (1952) readability index, Flesch’s (1948) readability index

Bridging Linguistic Gaps: Developing a Greek Text Simplification Dataset

Text simplification is crucial in bridging the comprehension gap in today’s information-rich environment. Despite advancements in English text simplification, languages with intricate grammatical structures, such as Greek, often remain under-explored. The complexity of Greek grammar, characterized by its flexible syntactic ordering, presents unique challenges that hinder comprehension for native speakers, learners, tourists, and international students. This paper introduces a comprehensive dataset for Greek text simplification, containing over 7500 sentences across diverse topics such as history, science, and culture, tailored to address these challenges. We outline the methodology for compiling this dataset, including a collection of texts from Greek Wikipedia, their annotation with simplified versions, and the establishment of robust evaluation metrics. Additionally, the paper details the implementation of quality control measures and the application of machine learning techniques to analyze text complexity. Our experimental results demonstrate the dataset’s initial effectiveness and potential in reducing linguistic barriers and enhancing communication, with initial machine learning models showing promising directions for future improvements in classifying text complexity. The development of this dataset marks a significant step toward improving accessibility and comprehension for a broad audience of Greek speakers and learners, fostering a more inclusive society.

Symbolic Analysis of the Bhagvad Gita: A Potential Methodology in the Indian Knowledge System Classroom Teaching

With the recent initiative to incorporate Indian Knowledge System into the university curricula at the undergraduate level across India, the Bhagvad Gita is being taught as part of this broader educational reform. Though the integration of the sacred text in the university curricula offers a unique opportunity to promote the revival of India’s rich cultural heritage, it cannot be denied that the text's density and complexity, with its intricate philosophical and metaphysical concepts, presents significant teaching challenges. The research paper offers symbolic analysis as a pedagogical approach to make the text simpler, more accessible and engaging in the classroom, helping students appreciate its depth and relevance. The paper suggests that Northrop Frye’s theory of symbols can be used as a pedagogical tool to discuss the layered meanings in the text. The paper explores the application of theory of symbols in classroom teaching, emphasizing a centripetal approach to reading texts that goes beyond surface interpretations to uncover deeper meanings. The methodology entails identification and analysis of symbols interwoven throughout the text in the classroom learning. It further discusses the process of locating the "centre" of the text, known as the monad, where all symbols converge. This central symbol transcends 'nature' and 'history' and serves as a focal point for understanding the text’s intricate and complex philosophical themes. By identifying this monad, educators can facilitate richer discussions and a more profound comprehension of literary works, thereby enriching students' analytical skills and appreciation of literature. The paper concludes that the pedagogical tool of Frye’s symbolic analysis in the class helps students recognize the text’s role as an ethical instrument for liberating the imaginative mind.

Text Complexity of Cambridge-delivered IELTS Academic Reading Tests: Comparability with IELTS Academic Reading Practice Tests from Other Publishers

Text Complexity of Cambridge-delivered IELTS Academic Reading Tests: Comparability with IELTS Academic Reading Practice Tests from Other Publishers

A Synchronic and Diachronic Study of Students’ Essays in Italian High Schools: Trends in Length, Complexity, and Referencing

In this work, we explore the use of digital technologies and statistical analysis to monitor how Italian secondary school students’ writing changes over time and how comparisons can be made across different high school types. We analyzed more than 2,000 exam essays written by Italian high school students over 13 years and in five different school types. Four indicators of writing characteristics were considered—text length, text complexity, and two indicators of source use, all extracted using natural language processing tools—which provided insights into students’ citation practices over time and in different school contexts. In particular, we measured the portion of students’ essays that included text from source material as well as the amount of copied text that was not properly referenced. We found that student essays became shorter in length over time while also getting more complex. We also found that the tendency to copy uncited text in the essay decreased. High school curricula predict different writing strategies: essays written by students attending scientific and humanistic high schools are longer and less subject to incorrect citations. We argue that such text analysis enables the study of writing features in high school classes and supports the evaluation of curricula.

Let’s Discover More API Relations: A Large Language Model-based AI Chain for Unsupervised API Relation Inference

APIs have intricate relations that can be described in text and represented as knowledge graphs to aid software engineering tasks. Existing relation extraction methods have limitations, such as limited API text corpus and affected by the characteristics of the input text. To address these limitations, we propose utilizing large language models (LLMs) (e.g., gpt-3.5) as a neural knowledge base for API relation inference. This approach leverages the entire Web used to pre-train LLMs as a knowledge base and is insensitive to the context and complexity of input texts. To ensure accurate inference, we design an AI chain consisting of three AI modules: API Fully Qualified Name (FQN) Parser, API Knowledge Extractor, and API Relation Decider. The accuracy of the API FQN Parser and API Relation Decider is 0.81 and 0.83, respectively. Using the generative capacity of the LLM and our approach’s inference capability, we achieve an average F1 value of 0.76 under the three datasets, significantly higher than the state-of-the-art method’s average F1 value of 0.40. Compared to the original CoT and modularized CoT methods, our AI chain design has improved the performance of API relation inference by 71% and 49%, respectively. Meanwhile, the prompt ensembling strategy enhances the performance of our approach by 32%. The API relations inferred by our method can be further organized into structured forms to provide support for other software engineering tasks.

Assessing effects of source text complexity on L2 learners’ interpreting performance: a dependency-based approach

Abstract Based on data from the English-Chinese interpretation corpus, we examined the relationship of source text complexity, captured using newly-developed dependency-based and traditional indices, to L2 learners’ interpreting performance captured using complexity, accuracy, and fluency. Best subsets regression and Poisson regression models yielded that the effectiveness of dependency-based indices including mean dependency distance, maximum dependency distance, dependency direction, and root distance, has been demonstrated across various dimensions of L2 learners’ performance. In contrast to the mixed results obtained from traditional indices, the consistent effect of dependency-based indices in these dimensions sheds light on the workings of cognitive processing. These findings provide preliminary support for the impact of dependency-based indices of source text on L2 learners’ interpreting performance, aiding in operationalizing task difficulty in L2 interpreting pedagogy. Moreover, they constitute product-based evidence for understanding bilingual switching.