Text Complexity Research Articles

Deep feature extraction with tri-channel textual feature map for text classification

The complexity and diversity of texts make it difficult for shallow text classification models to capture deeper text features. Therefore, this paper takes advantage of the BiLSTM-CNN hybrid network based on the self-attention mechanism to extract text features, and constructs a new text feature representation similar to RGB-channel images, that is, tri-channel text feature maps. Drawing on the effectiveness of ResNet and SA-Net in deep networks, we designed a deep feature extraction network to capture deeper features in the text. We use the tri-channel text feature map as the input to the deep feature extraction network, and propose the Deeper Feature Text Classification (DFTC) end-to-end text classification model. Experiments prove that the DFTC model is competitive compared to the most advanced methods on five challenging text classification datasets.

„Wyobrażam sobie albo pamiętam…”. Strategie narracyjne dokumentu historycznego na przykładzie filmu „Mała Zagłada” Natalii Korynckiej-Gruz

The article is an in-depth, contextual analysis of Natalia Koryncka-Gruz’s film A Minor Genocide. The author focuses on its narrative structure, in which the documentarian tried to reflect the nature of the work of memory and post-memory. The film is a documentary adaptation of Anna Janko’s (auto)biographical essay. The narrative complexity of the literary text is reflected in the film version, which takes the shape of a palimpsest. Its individual layers consist of verbal narratives from three generations of women, contemporary sequences, archival photographs and stylistically diverse animated sequences. The memory of traumatic experiences is superimposed on the idea of them. For the author, A Minor Genocide is an example of bringing Polish documentary film and literature closer together, as well as a new formula of a historical document.

ASSESSING USABILITY OF A MEDICATION ADHERENCE SYSTEM FOR PERSONS WITH MILD COGNITIVE IMPAIRMENT

Abstract Persons with mild cognitive impairment (PwMCI) are at risk of medication nonadherence due to prospective memory deficits. Over time, they are likely to experience accelerated deterioration in cognitive functions and may develop dementia when comorbid treatable health conditions, such as hypertension, are not well managed. A digital therapeutic system called Medication Education, Decision Support, Reminding, and Monitoring (MEDSReM©) has been designed for cognitively normal older adults to support hypertension medication adherence and is currently being tested in a randomized controlled trial. MEDSReM has the potential to support PwMCI with hypertension medication adherence, but it must be optimized specifically for this population. We conducted a cognitive walkthrough to identify usability issues and inform the redesign and optimization of the system for PwMCI. This usability technique is used to evaluate a system from a user’s perspective and is typically conducted by subject matter experts. We had 12 individuals with interdisciplinary expertise in cognitive aging, cognitive impairment, hypertension management, nursing, human factors, and health technology participate in the cognitive walkthrough. Usability issues related to figure-ground contrast, text complexity, and flexibility related to decision making were identified. Overall, recommendations included (1) simplifying the user interface (e.g., color contrast/textual cues), (2) eliminating options related to medication taking actions (e.g., snooze, take later) to mitigate memory-related errors, and (3) providing more direct, prescriptive instructions to minimize decision-making demands. Based on these recommendations, MEDSReM is currently being redesigned for testing with PwMCI. Our approach can inform the design of other mobile health tools.

Text content variables as a function of comprehension: Propositional discourse analysis

Text complexity impact on immediate recalls and range of metadiscourse markers remains a research niche due to the lack of multidisciplinary data necessary to shed light on the issue. The current study aims to identify effects of text complexity and Russian-English discourse differences on immediate text-based recalls relating to the amount and type of the information reproduced. For the research purposes we engaged 94 native Russian speakers as respondents in a text-retelling task to explore the amount of propositions recalled from an opinion article and the range of discourse markers employed. The reading text and text-based recalls were contrasted on informative and linguistic levels. The informative complexity of the reading text was evaluated on the basis of propositional analysis, and the linguistic complexity was carried out on the basis of descriptive parameters (word and sentence length, proportion of long words), readability index, word complexity and range of metadiscourse markers. The study revealed that the complexity level of the reading text is a strong predictor of propositional recall. The comparative analysis indicated a slight decrease in metrics of descriptive parameters. We also revealed that high ability readers make a choice in favor of superordinate propositions recalling about 60% of them and losing over 70% of the subordinate propositions. They also tend to shift the metadiscourse patterns of the original text from interactive to more logical ones by loosing hedges, emphatics and evidentials. The study furthers our understanding of cross-linguistic differences in the use of metadiscourse, its results will find application in discourse complexology and natural language processing.

Textual disclosure complexity and analysts’ weighting of information

This study investigates the influence of textual complexity in firms’ annual reports (Form 10-K) on financial analysts’ weighting of information. Drawing on a sample of U.S.-listed companies spanning from 1994 through 2020, we demonstrate that analysts tend to under-weight private information when 10-K reports are more readable, especially when the disclosed events indicate positive news, and when analysts face high workloads. Additionally, we find that analysts can identify managers’ opportunistic disclosure practices and rationally allocate weight between private and public information. Our findings suggest that analysts are affected by the underlying reporting and business complexity but not misled by the discretionary readability of 10-K reports when weighting information.

A data-driven global observatory addressing worldwide challenges through text mining and complex data visualisation.

Observing the world on a global scale can help us understand better the context of problems that engage us all. In this paper, we propose a data-driven global observatory methodology that puts together the different perspectives of media, science, statistics and sensing over heterogeneous data sources and text mining algorithms. We also discuss the implementation of this global observatory in the context of epidemic intelligence, monitoring the impact of the COVID-19 pandemic, and in the context of climate change, with a specific focus on water resource management. Moreover, we discuss the value of this global solution in local contexts and priorities, based on the exchange with stakeholders in municipalities, utilities and governmental institutions.

Features of lexical complexity: insights from L1 and L2 speakers

We discover sizable differences between the lexical complexity assignments of first language (L1) and second language (L2) English speakers. The complexity assignments of 940 shared tokens without context were extracted and compared from three lexical complexity prediction (LCP) datasets: the CompLex dataset, the Word Complexity Lexicon, and the CERF-J wordlist. It was found that word frequency, length, syllable count, familiarity, and prevalence as well as a number of derivations had a greater effect on perceived lexical complexity for L2 English speakers than they did for L1 English speakers. We explain these findings in connection to several theories from applied linguistics and then use these findings to inform a binary classifier that is trained to distinguish between spelling errors made by L1 and L2 English speakers. Our results indicate that several of our findings are generalizable. Differences in perceived lexical complexity are shown to be useful in the automatic identification of problematic words for these differing target populations. This gives support to the development of personalized lexical complexity prediction and text simplification systems.

The Mukokuseki Strategy and the Application of Pivot Translation in the Localization of Japanese Games

This article investigates the impact of using pivot language on the quality of game localization via a case study on the Japanese game Dark Souls 3. It first discusses the indirectness of the source text and the complexity of the target text in game localization by highlighting the “born translated” nature of the game narrative design. It then explicates the intricacy of having game audio voiced over in English only and including both direct and indirect translations in the localization workflow. It demonstrates the complication in evaluating localization quality from players’ perspectives and suggests that using English as a pivot language has certain advantages in ensuring multimodal cohesiveness in game localization. The rationale behind the localization approach is then explored through the lens of the mukokuseki strategy, which suggests that the application of pivot translation is driven by market globalization and, ultimately, the pursuit of economic gain.

Information Quality and Readability: ChatGPT's Responses to the Most Common Questions About Spinal Cord Injury

This study aimed to assess the quality, readability, and comprehension of texts generated by ChatGPT in response to commonly asked questions about spinal cord injury (SCI). The study utilized Google Trends to identify the most frequently searched keywords related to SCI. The identified keywords were sequentially inputted into ChatGPT, and the resulting responses were assessed for quality using the Ensuring Quality Information for Patients (EQIP) tool. The readability of the texts was analyzed using the Flesch-Kincaid grade level and the Flesch-Kincaid reading ease parameters. The mean EQIP score of the texts was determined to be 43.02±6.37, the Flesch-Kincaid reading ease score to be 26.24±13.81, and the Flesch-Kincaid grade level was determined to be 14.84±1.79. The analysis revealed significant concerns regarding the quality of texts generated by ChatGPT, indicating serious problems with readability and comprehension. The mean EQIP score was low, suggesting a need for improvement in the accuracy and reliability of the information provided. The Flesch-Kincaid grade level indicated a high linguistic complexity, requiring a level of education equivalent to approximately 14 to 15years of formal education for comprehension. The results of this study show heightened complexity in ChatGPT-generated SCI texts, surpassing optimal health communication readability. ChatGPT currently cannot substitute comprehensive medical consultations. Enhancing text quality could be attainable through dependence on credible sources, the establishment of a scientific board, and collaboration with expert teams. Addressing these concerns could improve text accessibility, empowering patients and facilitating informed decision-making in SCI.

Tranzitusok, exitusok és modulációk

The networked narrative complexities of Bodor texts are achieved by variations and iterations, and this is especially the case with the intricately written Sinistra körzet. This allows for the kind of alinear reading that makes possible the cross-referencing of the different textual levels (e.g. a plot point, a description, or a phrase uttered by a character) as well as the self-reflexive correlation of textual elements and speech events that otherwise appear in separate contexts. This means that the text opens itself up to additional semantic possibilities and self-representational shi”fting. The character and landscape descriptions in Bodor oft”en have the marked effect of blurring the lines between humans, animals, plants, and objects, thereby creating a rich field of metaphorical relations and modulations that venture into the ironic and the grotesque. Humans are substituted/reattributed as animalistic via the act of naming, in order to garner a sense of uniqueness, of the unknown, of the alien, but at the same time this substitution brings about both the annihilation and the familiarization of what is human. is paper explores the textual semiosis of Bodor texts through close readings, focusing on characterization (e.g. “vörös kakas”) and on soundscape descriptions which include the mixing of biophonic and antropophonic sounds. e narrative is “unreliable” and the speech events are idiomatic, resulting in a performative textual worlding and a grotesque-absurd modality that never quite become thetical. In Bodor, the transitions/transformations resist being allegorized.

Usability enhancement model for unstructured text in big data

The task of insights extraction from unstructured text poses significant challenges for big data analytics because it contains subjective intentions, different contextual perspectives, and information about the surrounding real world. The technical and conceptual complexities of unstructured text degrade its usability for analytics. Unlike structured data, the existing literature lacks solutions to address the usability of unstructured text big data. A usability enhancement model has been developed to address this research gap, incorporating various usability dimensions, determinants, and rules as key components. This paper adopted Delphi technique to validate the usability enhancement model to ensure its correctness, confidentiality, and reliability. The primary goal of model validation is to assess the external validity and suitability of the model through domain experts and professionals. Therefore, the subject matter experts of industry and academia from different countries were invited to this Delphi, which provides more reliable and extensive opinions. A multistep iterative process of Knowledge Resource Nomination Worksheet (KRNW) has been adopted for expert identification and selection. Average Percent of Majority Opinions (APMO) method has been used to produce the cut-off rate to determine the consensus achievement. The consensus was not achieved after the first round of Delphi, whereas APMO cut-off rate was 70.9%. The model has been improved based on the opinions of 10 subject matter experts. After second round, the analysis has shown majority agreement for the revised model and consensus achievement for all improvements that validate the improved usability enhancement model. The final proposed model provides a systematic and structured approach to enhance the usability of unstructured text big data. The outcome of the research is significant for researchers and data analysts.

Toward an Educative EEG-Based neuroIIR System for Adapting Contents

The pursuit of information is an integral aspect of a learner’s daily routine, often facilitated by information retrieval systems. In this study, we explore the correlation between text complexity and reading comprehension through the lens of neuroinformatics. Our objective is to ascertain whether it is feasible to infer the difficulty level of a text for a learner based on their brain activity, quantified by an electroencephalogram (EEG), during the information retrieval process. To accomplish this, we administered a section of the Woodcock Reading Mastery Test to our participants, evaluating their average reading fluency. Subsequently, the 18 participants under observation read paragraphs of varying complexity and responded to questions pertaining to the text. Utilizing the amassed data, we trained a deep learning model known as EEGNet to autonomously discern the complexity of the text being read, relying on the EEG signals. Furthermore, we conducted a comparison with previous research that employed a distinct characterization approach and employed a random forest algorithm for classification. The outcomes revealed that the present approach (EEGNet) outperformed the former method, achieving an accuracy rate of 80.83% across all subjects, whereas the random forest algorithm achieved a mere 50.28% accuracy. Our findings suggest the potential to classify the EEG signals in accordance with the level of difficulty a student encounters when comprehending the text. We propose several avenues for extending our work into real-time scenarios.

Kernel Proposal Network for Arbitrary Shape Text Detection.

Segmentation-based methods have achieved great success for arbitrary shape text detection. However, separating neighboring text instances is still one of the most challenging problems due to the complexity of texts in scene images. In this article, we propose an innovative kernel proposal network (dubbed KPN) for arbitrary shape text detection. The proposed KPN can separate neighboring text instances by classifying different texts into instance-independent feature maps, meanwhile avoiding the complex aggregation process existing in segmentation-based arbitrary shape text detection methods. To be concrete, our KPN will predict a Gaussian center map for each text image, which will be used to extract a series of candidate kernel proposals (i.e., dynamic convolution kernel) from the embedding feature maps according to their corresponding keypoint positions. To enforce the independence between kernel proposals, we propose a novel orthogonal learning loss (OLL) via orthogonal constraints. Specifically, our kernel proposals contain important self-information learned by network and location information by position embedding. Finally, kernel proposals will individually convolve all embedding feature maps for generating individual embedded maps of text instances. In this way, our KPN can effectively separate neighboring text instances and improve the robustness against unclear boundaries. To the best of our knowledge, our work is the first to introduce the dynamic convolution kernel strategy to efficiently and effectively tackle the adhesion problem of neighboring text instances in text detection. Experimental results on challenging datasets verify the impressive performance and efficiency of our method. The code and model are available at https://github.com/GXYM/KPN.

Research on Safety Risk Transfer in Subway Shield Construction Based on Text Mining and Complex Networks

Subway construction is often in a complex natural and human-machine operating environment, and that complicated setting leads to subway construction being more prone to safety accidents, which can cause substantial casualties and monetary losses. Thus, it is necessary to investigate the safety risks of subway construction. The existing literature on the identification and assessment of subway construction safety risks (SCSR) is susceptible to the influence of subjective factors. Moreover, although existing studies have explored the interrelationships between different risks, these studies usually analyze the interrelationships of single risks, lack the study of risk chain transfer relationships, and fail to find out the key path of risk transfer. Therefore, this paper innovatively combines text mining, association rules, and complex networks to deep mine subway construction safety incident reports and explore the risk transfer process. Firstly, it uses text mining technology to identify subway construction safety risks. Then, association rules are introduced to explore the causal relationships among safety risks. Finally, the key safety risks and important transfer paths of subway construction safety accidents (SCSA) are obtained based on the complex network model. Research results show that (a) improper safety management, unimplemented safety subject responsibilities, violation of operation rules, non-perfect safety responsibilities system and insufficient safety education and training are the key safety risks in SCSA; (b) two shorter key risk transfer paths in the subway construction safety network can be obtained: insufficient safety education and training→lower safety awareness→violation of operation rules→safety accidents; insufficient safety checks or hidden trouble investigations→violation of operation rules→safety accidents; (c) in the process of risk transfer, the risk can be controlled by controlling the key safety risk or cutting off the transfer paths. This paper provides new ideas and methods for SCSR identification and influence element mining, and the results of the study help safety managers propose accurate subway construction safety risk control measures.

FRAMING ANALYSIS OF CHINA'S RELATIONSHIP ABOUT THE SOUTH CHINA SEA DISPUTE ON CHANNEL NEWS ASIA

Channel News Asia has been an excellent news site in Asia and other foreign countries because of its exceptional phrase complexity and informative text which is at par to The Guardian or VICE, despite being CNN tier news site by the web and operational design. Currently, the South China Sea conflict is part of the CNA’s subtle information delivery regarding UC-China rivalry without sounding like Yellow Journalism. Using Robert N. Entman’s framing analysis, the author was able to discover how CNA delivers the geopolitical issues to provide information regarding US-China rivalry in trade market.

We Ask Students What they Understand, Not How they Understand: Making Reasoning Comprehension Processes Visible and Explicit

AbstractThis paper pushes beyond current debates over what is being called the science of reading to articulate a multidimensional complex conception of what is entailed in reading comprehension. Reading comprehension entails not only cognitive processes, but equally important is how issues of identity along multiple dimensions, perceptions of tasks and settings, social and emotional relationships intersect to shape effort and motivation to comprehend. The paper argues that a central problem in the typical teaching of reading comprehension in the K–12 sector is insufficient attention to processes or meaning making. Typical instruction focuses on the outcomes of comprehension, but not on how readers can/should go about trying to make sense of texts. Such instruction can be created by communities of teachers who examine the sources of text difficulty in texts they are teaching. The paper offers explicit criteria for evaluating sources of text complexity in informational and literary texts, and in so doing push beyond the boundaries of traditional Lexile scores.

Complex Syntax Production in Informational Writing by Students With Language Impairment From Diverse Linguistic Backgrounds

The primary objective of this study was to compare the syntactic complexity of informational texts produced by fifth-grade students (a) with and without language impairment and (b) with and without native English-speaking proficiency on a curriculum-based reading and writing task. Expository writing samples produced by 114 children enrolled in the fifth grade were analyzed at the utterance level for five features of complex syntax, including the frequency of utterances containing complex syntax and specific clause types (conjoined, subordinate, relative, full complement). For each of the four groups, we report frequency counts, means, standard deviations, and ranges of performance across the five syntax measures. Multivariate analysis of variance revealed there were significant group differences on measures. Specifically, children with typical language development, regardless of English proficiency level, wrote more words, utterances, and different word roots than their peers with language impairment. When productivity (i.e., text length) in the writing samples was controlled, multivariate analysis of variance revealed there was a significant difference between groups in use of relative clauses, but not for the use of conjoined, subordinate, or full complement clauses. In particular, English proficient students with language impairment produced a greater proportion of utterances with relative clauses. A post hoc correlation analysis showed moderate, positive correlations among writing cohesion and variables of complex syntax. We consider multiple implications for clinical practice and further research.

From Suspicion to Sincerity in Composition Pedagogy

Abstract Recent advocates of postcritique urge scholars not to read texts suspiciously but instead to regard texts as capable of saying what they mean and, accordingly, to take those meanings seriously. While a suspicious disposition underlies much of introductory composition pedagogy, especially the teaching of argument, postcritique has made little entry into discourses of undergraduate instruction. Attending to the New Sincerity movement in American literature, film, and music after 1980, this essay examines how teaching texts that emphasize their own sincerity (and the difficulty of achieving sincere expression) can encourage students to regard argument and interpretation not as suspicious practices but as means for a generous mode of description that does not sacrifice the complexity of a given text.

Topic Modeling for Text Structure Assessment: The case of Russian Academic Texts

Background: Automatic assessment of text complexity levels is viewed as an important task, primarily in education. The existing methods of computing text complexity employ simple surface text properties neglecting complexity of text content and structure. The current paradigm of complexity studies can no longer keep up with the challenges of automatic evaluation of text structure.
 Purpose: The aim of the paper is twofold: (1) it introduces a new notion, i.e. complexity of a text topical structure which we define as a quantifiable measure and combination of four parameters, i.e. number of topics, topic coherence, topic distribution, and topic weight. We hypothesize that these parameters are dependent variables of text complexity and aligned with the grade level; (2) the paper is also aimed at justifying applicability of the recently developed methods of topic modeling to measuring complexity of a text topical structure.
 Method: To test this hypothesis, we use Russian Academic Corpus comprising school textbooks, texts of Russian as a foreign language and fiction texts recommended for reading in different grades, and employ it in three versions: (i) Full Texts Corpus, (ii) Corpus of Segments, (iii) Corpus of Paragraphs. The software tools we implement include LDA (Latent Dirichlet Allocation), OnlineLDA and Additive Regularization Of Topic Models with Word2vec-based metric and Normalized Pairwise Mutual Information.
 Results: Our findings include the following: the optimal number of topics in educational texts varies around 20; topic coherence and topic distribution are identified to be functions of grade level complexity; text complexity is suggested to be estimated with structural organization parameters and viewed as a new algorithm complementing the classical approach of text complexity assessment based on linguistic features.
 Conclusion: The results reported and discussed in the article strongly suggest that the theoretical framework and the analytic algorithms used in the study might be fruitfully applied in education and provide a basis for assessing complexity of academic texts.

Cognitive complexity measures for educational texts: Empirical validation of linguistic parameters

The article presents a study conducted within the framework of discourse complexology - an integral scientific domain that has united linguists, cognitive scientists, psychologists and programmers dealing with the problems of discourse complexity. The issue of cognitive complexity of texts is one of the central issues in discourse complexology. The paper presents the results of the study aimed to identify and empirically validate a list of educational texts’ complexity predictors. The study aims to identify discriminant linguistic parameters sufficient to assess cognitive complexity of educational texts. We view text cognitive complexity as a construct, based on the amount of presented information and the success of reader-text interactions. The idea behind the research is that text cognitive complexity notably increases across middle and high schools. The research dataset comprises eight biology textbooks with the total size of 219,319 tokens. Metrics of text linguistic features were estimated with the help of automatic analyzer RuLingva (rulingva.kpfu.ru). Linguistic and statistical analysis confirmed the hypothesis that text syntactic and lexical parameters are discriminative enough to classify different levels of cognitive complexity of educational texts used in middle and high schools. Text parameters that manifest variance in cognitive complexity include lexical diversity (TTR); local argument overlap; abstractness index; number of polysyllabic words, Flesch-Kincaid Grade Level; number of nouns and number of adjectives per sentence. Empirical evidence indicates that the proposed approach outperforms existing methods of text complexity assessment. The research results can be implemented in the system of scientific and educational content expertise for Russian school textbooks. They can also be of some use in the development of educational resources and further research in the field of text complexity.