Textual Content Research Articles

Pragmatic particles as distinguishing features of registers

Abstract The study explores the potential of utilizing particle use data to differentiate between seven Estonian registers: everyday conversation, institutional interaction, printed media (newspapers), prose fiction, academic prose, instant messaging, and internet comments. The objective is to develop a simple yet effective model that enables researchers to comprehend the internal logic behind register differentiation based on particle use. Particles are considered promising differentiators due to their independence from text content. The article outlines the chosen method, the model creation, and the testing process. A key finding reveals that hierarchical relationships between particles within registers prove more reliable indicators than general use frequencies. The method involves establishing correspondences between particle pairs and register pairs, facilitating the measurement of distances between registers. During testing, the model demonstrates high accuracy across registers, encountering some difficulties in categorizing fiction and institutional interaction. Overall, the study confirms the efficacy of the proposed method in distinguishing registers based on particle use, underscoring the significance of particles in linguistic analysis.

Borders and Bridges in Gender Discourse: A Comparative Study of Feminist Stand-up Comedy in China and the United States

This paper selects the textual content of Yang Li’s and Iliza’s stand-up comedy for a comparative analysis, aiming to foster a dialogue between Chinese and American feminist stand-up comedy. This research findings uncover numerous similarities as well as several noteworthy differences. Additionally, to further explore the female discourse in Chinese and American feminist stand-up comedy, this paper employed quantitative analysis to assess the audience sentiment in the comment sections of the two comedians’ videos, aiming to analyze the discourse power and expression rights of these two feminist comedians. The findings offer insight into the potential and difficulties faced by female comedians tackling gender problems by highlighting the cultural quirks and social attitudes toward feminist humor in both nations. In addition, the study looks at how these comedians’ performances affect the public conversation about feminism and gender equality, offering insights into how feminist comedy is changing in China and the US.

Analysis and Recommendations on Empowering Non-Art Background College Students to Generate Visual and Textual Content Using AIGC Tools

This study selects students from Shenzhen Polytechnic University with non-art backgrounds as the sample group. By integrating questionnaire surveys with specific case studies, it explores how AIGC tools empower these students to generate both visual and textual content and design related works. However, their usage primarily concentrates on basic functions such as generating texts and images. It is therefore recommended that systematic courses be developed to guide non-art background students in using AIGC tools more effectively and comprehensively, enabling them to perform more complex tasks.

Hearing Rhythm, Seeing Rhythm: A Research Approach to Reimagining Traditions of Chinese Poetic ‘rules of rhythm’

Throughout the history of China, from early dynasties into contemporary times, aspects of poetry, painting, and writing have been grounded in varied but deeply connected ‘rules of rhythm’. These ‘rules’ permeate in diverse yet identifiable and recognizable forms. The research in this project is aimed at both adopting and adapting these rules of rhythm or rhythmic patterns in an experimental form that reimagines a tradition of ‘lyric aesthetics’ in the context of the contemporary interplay between sound, image, and writing. In this sense, the investigation is at once one of ‘reclamation’ and extension. The basic ‘question’ in these experiments is whether it is possible to represent sonic patterns themselves — that is, by both visualizing sound structures and ‘sonifying’ visual forms (irrespective of textual content). The text focuses on the historical establishment of methodological approaches (as a way to identify their forms) and then transforming/translating these in combinations made possible by blending the sonic and visual. As a particular attribute of Chinese expression, these ‘rules’ or forms are cultural traits that have endured through many centuries and challenges and, more importantly in this case, can be (re)iterated in dynamic, fluid, and revitalized contemporary artistic configurations that aim at seeing sound, writing sound, and performing rhythm. <style> /* rules to make button only show up in META */ .download-accessible { display:none; } .meta-right-col .download-accessible { display: inline-block; padding: 9px; margin-bottom:25px; border: 1px solid black; background-color:white; } </style> <a class="download-accessible" href="download-media?work=3155507&file=3155517" title="This accessible PDF is a derivative of the original which it is meant to support and not replace.">Download Accessible PDF</a>

The union of linguistics and prompt engineering: linguistic features of prompts to neural networks

This article offers a linguistic perspective on the study of the interaction between humans and Artificial Intelligence. The focus of the work is on analyzing human communication with neural networks which generate textual content. With the advent of text-based interfaces for neural networks, where programming skills are no longer required for human-AI communication, and it is sufficient to create text queries (prompts), one of the key skills for prompt engineers (specialists in prompt creation) has become the ability to ‘program’ and obtain AI responses corresponding to the query while using natural language resources. This work is aimed at studying, identifying, and describing the linguistic features of prompts and their parameters that influence the stylistic correctness of the generated responses by the neural networks.

Determinants of receiving government support by small and medium-sized enterprises in Russia

Government support measures help companies overcome various restrictions impeding their growth and develop ment, which is of special relevance for small and medium-sized enterprises (SMEs). These measures, therefore, must reach the target companies and ensure efficient allocation of resources. The article reveals key characteristics of SMEs in Russia that increase the probability of them receiving government support. The research methodology is based on public interest theory, public choice theory and the concept of economic freedom. The research method is a two-step regression analysis, in which the binary choice model is first evaluated, and then the LASSO model is used to assess the impact of the textual content of the companies’ websites. The dataset of the study is represented by data on 2,176 Russian metallurgical enterprises for 2020–2023, of which 57% were receiving government support. The results of the first step of evaluation showed that the organization’s age and size meas ured by revenue were the key factors associated with the probability of receiving government support. At the second step, we found that the mention of certain goods or materials on the websites, the enterprise’s location and orientation towards foreign stakeholders were significantly related to the likelihood of getting state support. The study offers tools for analyzing the determi nants of government support for SMEs and highlights the importance of this analysis for more efficient allocation of government resources.

Systemic Functional Linguistic Based on Semantic Web

This research aims to combine a functional and systematic linguistic approach with Semantic Web technology. The main aim of this research is to develop a framework that can facilitate systematized analysis and language modeling in the context of the semantic web. The research method used is a qualitative approach with focused data analysis. The data collected will involve text and web content relevant to the Semantic Web framework. The analysis was carried out using the principles and tools of systematic functional linguistic analysis to identify language structure, communicative function, and meaning in web content. The results of this research are expected to provide new insights into the understanding of how language can be organized and analyzed regularly in the context of the semantic web. Apart from that, this research is also expected to contribute to the development of Semantic Web technology by using a functional linguistic approach. It is hoped that the conclusions of this research will provide a deeper understanding of the role of language in the context of the semantic web and provide a basis for the development of more sophisticated and effective Semantic Web-based applications and systems.

Enhancing domain-specific text generation for power grid maintenance with P2FT.

The digitization of operation and maintenance in the intelligent power grid equipment relies on a diverse array of information for smart decision-making. In the domain of intelligent decision generation, proficiency is contingent upon extensive learning from copious amounts of text. This necessitates not only robust processing capabilities but also a high level of specialization. In addressing situations where authorization is lacking, pre-trained language models (PLMs) have already provided ideas when confronted with specialized domains or tasks. In consideration of the complexity of textual content in the field of the power grid, which encompasses a multitude of specialized knowledge and involves an abundance of proprietary terminology, we have undertaken an exploration of pre-trained model specialization using the power grid domain as an example, specifically for the task of generating maintenance strategies. A two-stage fine-tuning approach (P2FT) is employed, utilizing a large-scale pre-training model specifically designed for natural language processing. The efficacy and practical value of this method were evaluated through multiple metrics, juxtaposed with other advanced approaches involving low-parameter or parameter-free fine-tuning methods. Through a meticulous analysis and validation of experimental outcomes, we have corroborated the feasibility and practical application value of employing this approach for pre-trained model specialization. Additionally, it has furnished valuable guidance for text generation within both the Chinese language domain and the power grid domain.

Text Command Intelligent Understanding for Cybersecurity Testing

Research on named entity recognition (NER) and command-line generation for network security evaluation tools is relatively scarce, and no mature models for recognition or generation have been developed thus far. Therefore, in this study, the aim is to build a specialized corpus for network security evaluation tools by combining knowledge graphs and information entropy for automatic entity annotation. Additionally, a novel NER approach based on the KG-BERT-BiLSTM-CRF model is proposed. Compared to the traditional BERT-BiLSTM model, the KG-BERT-BiLSTM-CRF model demonstrates superior performance when applied to the specialized corpus of network security evaluation tools. The graph attention network (GAT) component effectively extracts relevant sequential content from datasets in the network security evaluation domain. The fusion layer then concatenates the feature sequences from the GAT and BiLSTM layers, enhancing the training process. Upon successful NER execution, in this study, the identified entities are mapped to pre-established command-line data for network security evaluation tools, achieving automatic conversion from textual content to evaluation commands. This process not only improves the efficiency and accuracy of command generation but also provides practical value for the development and optimization of network security evaluation tools. This approach enables the more precise automatic generation of evaluation commands tailored to specific security threats, thereby enhancing the timeliness and effectiveness of cybersecurity defenses.

How dark traits and regulatory focus manifest in online hotel reviews: a textual content analysis

ABSTRACT This study seeks to explore how Dark Triad of psychopathy, Machiavellianism, and narcissism interact with consumers’ regulatory focus (prevention-focus and promotion-focus) to influence online consumer reviews in the hotel industry. We conducted an empirical analysis of 1,185,940 hotel reviews from TripAdvisor.com, utilizing the textual content of online reviews and employing Language Inquiry and Word Count analytical tools. The research reveals a multifaceted link between regulatory focus and Dark Triad in online reviews. The results showed that individuals with a promotion focus tend to give higher ratings than those with a prevention focus. Dark Triad significantly influences how regulatory focus impacts online reviews. This study initiates a new academic dialogue on the interaction between regulatory focus and Dark Triad traits within online consumer behavior studies. It provides practical guidance for the hotel industry to refine their management and response strategies by acknowledging the impact of consumer personality traits on online reviews.

Kárpátalja egyházi fenntartású iskoláinak virtuális nyelvi tájképe

We live in an age where everywhere we look, thousands of pieces of information are streaming towards us. This fact in itself highlights the importance of the linguistic landscape in our everyday lives, as everything that surrounds us, everything we perceive influences our thinking and therefore our use of language. In the 21st century, we need to pay particular attention to the virtual world, as a large percentage of the younger and even older generations receive a significant proportion of their stimuli through some types of electronic communication tools or social platforms. Schools have a particularly important role in language development and formation, as young people, the future generation, are most shaped within their walls. In the present study, we analysed a sample of posts published in the last few years on the social platforms (official website and Facebook) of the ecclesiastical schools of Transcarpathia. We relied on statistical and comparative analysis of the textual content of these entries. In addition, we examined the meta-linguistic information of the entries. The overall visual appearance of the website was also taken into consideration, i.e., the types of images published were also investigated. In the research, it was assumed that all the official webpages of the schools under investigation contain entries in both Ukrainian and Hungarian. We also assessed whether a higher percentage of entries in Hungarian appeared on the schools’ online interfaces. Results showed that the number of interactions by people who followed the sites was higher for Hungarian language posts, advertisements and articles. We assumed that religious posts and symbols regularly appeared on the websites and Facebook pages, and that a higher percentage of these posts and symbols appeared in Hungarian. It was also hypothesized that the volume of texts in Ukrainian was smaller than the amount of texts in Hungarian. The research has confirmed that both Ukrainian and Hungarian entries are found on the websites, with a higher share of entries in Hungarian. It also confirmed our hypothesis that followers of the websites tend to respond more frequently to posts in Hungarian, due to the fact that the majority of followers use Hungarian as their mother tongue. Last but not least, the research revealed that religious posts and symbols regularly appear on the official websites and Facebook pages of the lyceums, and the language of these posts is mostly Hungarian.

Beyond Text: Multimodal Credibility Assessment Approaches for Online User-Generated Content

User-generated content (UGC) is increasingly becoming prevalent on various digital platforms. The content generated on social media, review forums, and question–answer platforms impacts a larger audience and influences their political, social, and other cognitive abilities. Traditional credibility assessment mechanisms involve assessing the credibility of the source and the text. However, with the increase in how user content can be generated and shared (audio, video, and images), multimodal representation of UGC has become increasingly popular. This article reviews the credibility assessment of UGC in various domains, particularly identifying fake news, suspicious profiles, and fake reviews and testimonials, focusing on both textual content and the source of the content creator. Next, the concept of multimodal credibility assessment is presented, which also includes audio, video, and images in addition to text. After that, the article presents a systematic review and comprehensive analysis of work done in the credibility assessment of UGC considering multimodal features. Additionally, the article provides extensive details on the publicly available multimodal datasets for the credibility assessment of UGC. In the end, the research gaps, challenges, and future directions in assessing the credibility of multimodal UGC are presented.

Kiswahili Translation in Digital Media Platforms: Bridging Language Gaps in the AI Age

Purpose: As digital media platforms continue to proliferate globally, the need for multilingual content in these platforms has become increasingly apparent. Translation of online content is a necessity for netizens and other digital platform users who seek knowledge and information around a myriad of user-generated content on diverse issues and topics. Being a widely spoken language that goes beyond the borders of East Africa, Kiswahili’s significance in bridging linguistic barriers and facilitating communication and information access for Kiswahili speakers is clear. The purpose of this exploratory research paper was to examine the state of Kiswahili as a language of translation through different technologies across varied translation apps and social media platforms such as Facebook, You Tube, Tik Tok and Netflix. Methodology: The study adopted a descriptive research design. Data collection was done through online data mining tools including API access for web scrapping and manual extraction of content through random search engine searches. The data for the study included 10 sentences, 10 videos and 25 specific word forms such as proverbs, spiritual/religious words, similes and cultural words for food, beauty and family ties. Content analysis for both text and multimedia content such as written sentences, social media posts and videos was done. The study analyzed translation outputs of 14 Kiswahili translation apps in order to establish the status and accuracy of these translations. Of key consideration was the role of AI technology in the translation apps and the effect it has on the translated texts across different apps. The challenges and emerging issues facing Kiswahili translation through digital technologies were also explored. Findings: The findings of this study showed that the Apps used for Kiswahili translation have demonstrated great translation accuracy in basic semantic forms, with an average score of 60% accuracy upon random searches and between 10-30% accuracy for cultural forms and idiomatic expressions. Unique Contribution to Theory, Practice and Policy: The study was guided by the Skopos theory by Hans Vermeer (1978) which was complemented by the Relevance Theory by Dan Sperber and Deidre Wilson (1986). These Apps seem to have the feature of interoperability in that they seem to produce the same results suggesting that the AI supported feature in the apps may be interconnecting the apps during translation. The study advocates for more AI integration translation apps as this will help standardize Kiswahili translation corpus.

Improved bidirectional attention flow (BIDAF) model for Arabic machine reading comprehension

Abstract Machine reading comprehension (MRC) refers to the process of instructing machines to comprehend and respond to inquiries based on a provided text. There are two primary methodologies for achieving this: extracting answers directly from the text or predicting them. Extracting answers involves anticipating the specific segment of text containing the answer, pinpointed by its starting and ending indices within the paragraph. Despite the increasing interest in MRC, exploration within the framework of the Arabic language faces limitations due to various challenges. A significant impediment arises from the inadequacy of resources available for Arabic textual content, which impedes the development of effective models. Furthermore, the inherent intricacies of Arabic, manifesting in its diverse linguistic forms including classical, modern standard, and colloquial, present distinctive hurdles for tasks involving language comprehension. This paper proposes an enhanced version of the bidirectional attention flow (BIDAF) model for Arabic MRC, constructed upon the Arabic Span-Extraction-based Reading Comprehension Benchmark (ASER). ASER comprises 10,000 sets of questions, answers, and passages, partitioned into a training set constituting 90% of the data and a testing set making up the remaining 10%. By introducing a new input feature based on parts-of-speech (POS) word embeddings and replacing Bidirectional Long Short-Term Memory (bi-LSTM) with bidirectional gated recurrent unit, significant improvements were observed. Eight different POS word embeddings were generated using both Continuous Bag of Words (CBOW) and Skip-gram methods, with varying dimensionalities. Evaluation metrics, including exact match (EM) and F1-measure, were utilized to assess model performance, with emphasis on the latter for its accuracy. The proposed enhanced BIDAF model achieved a remarkable accuracy of 75.22% on the ASER dataset, demonstrating its efficacy in Arabic MRC tasks. Additionally, rigorous statistical evaluation using a two-tailed paired samples t-test further validated the findings, highlighting the significance of the proposed enhancements in advancing Arabic language processing capabilities.

Analysis of feature extraction techniques for sentiment analysis of tweets

Over the past few years, sentiment analysis has moved from social networking services like LinkedIn, Facebook, YouTube, Twitter, and online product-based reviews to determine public opinion or emotion using social media textual contents. The methodology includes data selection, text pre-processing, feature extraction, classification model, and result analysis. Text pre-processing is an important stage in structuring data for improved performance of our methodology. The feature extraction technique (FET) is a crucial step in sentiment analysis as it is difficult to obtain effective and useful information from highly unstructured social media data. A number of feature extraction techniques are available to extract useful features. In this work, popular feature extraction techniques including bag of words (BOW), term frequency and inverse document frequency (TF-IDF), and Word2vec are compared and analyzed for the sentiment analysis of social media contents. A method is proposed for processing text data from social media networks for sentiment analysis that uses support vector machine as a classifier. The experiments are carried on three datasets of different context namely US Airline, Movie Review, and News from Twitter. The results show that TF-IDF consistently outperformed other techniques with best accuracy of 82.33%, 92.31%, and 99.10% for Airline, Movie Review, and News datasets respectively. It is also found that the proposed method performed better than some existing methods.

Identifying AI-Generated Research Papers: Methods and Considerations

Recent advancements in natural language generation (NLG) have revolutionized content creation, enabling artificial intelligence (AI) tools to produce coherent and seemingly authentic texts, including scholarly papers. While AI-generated content offers efficiencies in speed and volume, concerns over authenticity, ethical implications, and academic integrity persist. This review explores methods and considerations for identifying AI-generated research papers, emphasizing the need to distinguish between human-authored and AI-generated content to uphold scholarly standards and ensure transparency in research. Key detection techniques include textual analysis, metadata examination, and content evaluation. Ethical concerns regarding AI's role in research are also discussed, underscoring the importance of ongoing research to refine identification methodologies and maintain research integrity.

Exploring Twitter discourse with BERTopic: topic modeling of tweets related to the major German parties during the 2021 German federal election

AbstractWe present a study in the context of computational social science that explores the topics debated in the context of the 2021 German Federal Election by using the topic modeling technique BERTopic. The corpus consists of German language tweets posted by political party accounts of the major German parties, as well as tweets by the general public mentioning the party accounts. We examined the textual content of the tweets but also included the text in images that were posted into the analysis by extracting the text using optical character recognition (OCR). Our results show that the most frequently discussed topics are party-oriented policies (including call-to-action content), climate policy and financial policy, with these topics being discussed in tweets by both, the political party accounts and tweets by accounts mentioning them. In addition, we observed that some topics were discussed consistently throughout the year, such as the COVID-19 pandemic, climate policy or digitization, while other topics, such as the return to power of the Taliban in Afghanistan or Israel were debated to a greater extent at limited time frames during the election year.

Enhancing integration testing efficiency through AI-driven combined structural and textual class coupling metric

Integration testing, a critical and resource-intensive phase in the software development lifecycle, can account for up to a high percentage of the total testing cost. Identifying classes with high coupling is crucial for efficient integration testing, as these classes are more susceptible to the impact of maintenance-related changes. This research introduces a novel metric called <i>combined structural and textual class coupling</i> (CSTCC), which harnesses the power of artificial intelligence (AI) techniques to predict and rank the most critical classes in an object-oriented software system. CSTCC integrates structural coupling metrics with <i>latent semantic indexing</i> (LSI)-based textual coupling, providing a comprehensive measure of class coupling. LSI, an information retrieval technique, analyses the semantic relationships between classes based on their textual content, enabling CSTCC to capture both structural and conceptual dependencies, resulting in a more accurate identification of high-risk classes. The effectiveness of the proposed approach is rigorously evaluated using mutation testing on four Java open-source projects, and the results demonstrate that test cases developed based on CSTCC achieve high mutation scores, indicating their ability to detect a significant percentage of integration faults. By focusing testing efforts on high-coupling classes identified by CSTCC, developers can potentially save time and cost during integration testing. The results demonstrate that test cases developed based on CSTCC achieve high mutation scores, ranging from 98% to 100%, indicating their ability to detect a significant percentage of integration faults. Additionally, the approach results in substantial efficiency gains, with a notable reduction in the number of test cases needed, saving up to 33.3% of the testing effort in some cases. By focusing testing efforts on high-coupling classes identified by CSTCC, developers can potentially save time and cost during integration testing. The CSTCC metric provides a novel and effective approach to prioritize testing resources and improve the efficiency of integration testing in object-oriented software systems.

Clinical Reasoning Unveiled: Demystifying Heuristics and Hermeneutics in Neurological Clinical Practice

Introduction: Medical practice, particularly in Neurology, is often influenced by cognitive biases that can compromise diagnostic and therapeutic accuracy, leading to significant patient consequences. Clinical reasoning strategies are fundamental, yet the complexity of these processes is frequently underestimated. Understanding these strategies is crucial to ensure accurate diagnoses and effective treatments, thereby reducing clinical errors and optimizing patient outcomes. Despite extensive literature on cognitive biases, there is a need for a systematic synthesis addressing the peculiarities of neurological clinical reasoning, considering heuristics (mental shortcuts or rules of thumb that simplify decision-making) and hermeneutics (the theory and methodology of interpretation, especially of textual and symbolic content). Objectives: To review and systematize the main clinical reasoning strategies in Neurology, highlighting common biases and proposing methods to improve diagnostic and therapeutic accuracy. Methodology: A comprehensive literature review focusing on familiarity biases, Dual Process Theory, and heuristics (anchoring, availability, and representativeness). Additionally, the application of hermeneutics in hypothesis formulation and the importance of understanding the patient as a complex system were analyzed. Results: Specialized neurologists often rely on heuristics, which means a cognitive shortcut. Recognizing this process can mitigate anchoring and availability biases. The hermeneutic approach, on the other hand, is essential for interpreting variables in the patient’s context, integrating historical and social aspects. Conclusion: Understanding clinical reasoning processes and systematically teaching these strategies are crucial to enhancing the efficiency and effectiveness of clinical decision-making, reducing diagnostic errors, and improving patient outcomes.

Improvement for deep learning for suicide and depression identification with unsupervised label correction

Abstract. Our study try to address the challenge of accurately identifying depression and suicidal ideation on social media platforms by introducing an enhanced novel methodology for unsupervised feature selection and label correction. Utilizing advanced word embedding models like BERT and the Universal Sentence Encoder, we transform textual content into dense numerical vectors that capture the nuanced emotional context of online discussions. Our approach enhances these embeddings with a deep neural network (DNN) to extract distinctive features, reducing the dimensionality of the data through Principal Component Analysis (PCA). For label correction, we employ clustering techniques including OPTICS, K-medoids, and hierarchical clustering, which are robust against noisy data points. We then train classifiers using CNN, DNN, logistic regression, and random forest algorithms, evaluated with metrics such as accuracy, precision, recall, F1 score, and AUC. This methodology improves the accuracy of classifying depressive and suicidal sentiments to some extent, assisting to utilize the vast data available on social media to advance mental health diagnostics and interventions.