Media Texts Research Articles

Position-context additive transformer-based model for classifying text data on social media

In recent years, the continuous increase in the growth of text data on social media has been a major reason to rely on the pre-training method to develop new text classification models specially transformer-based models that have proven worthwhile in most natural language processing tasks. This paper introduces a new Position-Context Additive transformer-based model (PCA model) that consists of two-phases to increase the accuracy of text classification tasks on social media. Phase I aims to develop a new way to extract text characteristics by paying attention to the position and context of each word in the input layer. This is done by integrating the improved word embedding method (the position) with the developed Bi-LSTM network to increase the focus on the connection of each word with the other words around it (the context). As for phase II, it focuses on the development of a transformer-based model based primarily on improving the additive attention mechanism. The PCA model has been tested for the implementation of the classification of health-related social media texts in 6 data sets. Results showed that performance accuracy was improved by an increase in F1-Score between 0.2 and 10.2% in five datasets compared to the best published results. On the other hand, the performance of PCA model was compared with three transformer-based models that proved high accuracy in classifying texts, and experiments also showed that PCA model overcame the other models in 4 datasets to achieve an improvement in F1-score between 0.1 and 2.1%. The results also led us to conclude a direct correlation between the volume of training data and the accuracy of performance as the increase in the volume of training data positively affects F1-Score improvement.

A Corpus-Based Approach in Vocabulary Research: Defining the Word of the Year 2023 in Kazakh

The Word of the Year (WOTY) is an event held in various countries and regions to determine the most relevant, significant, and popular words and expressions that reflect not only the linguistic but also the socio-cultural aspects of the country. This paper aims to identify the most frequently used words/phrases in Kazakh for 2023 to be nominated for the WOTY title. The research methods include media discourse analysis and quantitative analysis using a corpus-based approach. A computer program, #LancsBox 6.0, generated a dataset—a research corpus consisting of 500 texts published on Kazakh news platforms throughout 2023. The results indicated that: 1) the conjunction “jáne” [and] had the highest frequency and occurrence in the research corpus; 2) the extracted words with high frequency indicators might serve as candidates for WOTY 2023, such as “Kazakhstan”, “jana” [new], “jyly” [year], “kerek” [need], “jumys [work]”; 3) WOTY “artificial intelligence” named by other global sources showed a high frequency indicator in Kazakh media texts. The study contributed with the generated corpus of media texts in Kazakh for 2023. The significance of our study is highlighted by the pioneering linguistic assessment in Kazakh language, which involves the analysis of media discourse publications based on corpus outcomes.

Research on Media Text Translation Based on Information Retrieval

This paper focuses on the research of media knowledge text translation based on information retrieval, and discusses how information retrieval technology can affect and optimize the translation process and results of media texts. By combining the basic principles of information retrieval technology and its practical application scenarios, this paper analyzes the characteristics of media texts and their special requirements for translation. Through the design experiment, the effect of information retrieval technology on improving translation quality, efficiency and reaction cost is evaluated. The experimental results show that this technology can significantly improve the accuracy and fluency of translation, shorten the translation cycle and reduce the cost. The application prospect of information retrieval in the field of translation will be broader. This study can not only provide a new perspective and method for the translation of media texts, but also contribute to the improvement of cross-cultural communication and information dissemination efficiency.

Mother Tongue–Based Instruction in Pakistani Print Media: Implications for Language Policy and Epistemic Inclusion

ABSTRACTThere is a growing interest in the role of mass media in language policy‐making, as it not only promotes the views and ideologies of various policy stakeholders but also plays an active role in shaping language policy development. Although the print media writers' insight provides a valuable perspective that has language policy implications, no previous studies investigated Pakistani print media writers' orientations toward indigenous languages. To fill the gap, this qualitative study aims to explore how indigenous languages are conceptualized in the discourses of Pakistani print media texts. Data were collected from Pakistani print media texts. By using Ruíz's framework of “language orientations” as the analytical tool, this study found that they largely considered the use of the mother tongue as a medium of instruction and a resource in the Pakistani context. Specifically, the analysis highlights the educational, cultural, and emotional significance of integrating mother tongues into education. However, complexities arise; Pervez, a media writer, highlights that promoting Punjabi as a medium of instruction in Punjab may provoke resistance from speakers of minority languages like Saraiki and Potohari, underscoring the need for inclusive language policies that reflect linguistic diversity. Moreover, the study highlights that the emotional connection fostered by mother tongues is not explicitly acknowledged in Ruíz's language orientations framework. It suggests incorporating this dimension, as valuing regional languages through literature and education can reduce stigma, build linguistic pride, and support equitable language policies. This study also contributes by offering policy implications.

Shaping trust and tension: Strategic leaks and their impact on global cybersecurity norms

This study investigates the role and impact of strategic leaks in shaping cyber diplomacy between the United States and China. Using data from incident databases, public opinion surveys, cybersecurity agreements, and textual media sources spanning 2010 to the present, a robust quantitative methodology was employed. Techniques included regression analysis, interrupted time series, and Pearson correlation to evaluate diplomatic outcomes, trust dynamics, and ethical considerations. Findings reveal that events like the Microsoft Exchange Hack exerted the most significant impact, with a 0.90 six-month influence score, and catalyzed multilateral cybersecurity collaborations. Ethical concerns such as misattribution showed a negative correlation with public trust (r=-0.75) but positively influenced support for transparency (r=0.70). Recommendations include fostering multilateral coalitions, developing frameworks for managing leaks, and promoting balanced narratives to align escalation strategies with long-term goals. These insights provide actionable pathways for addressing the dual-edged nature of leaks in international cybersecurity governance. Keywords: Strategic Leaks, Cyber Diplomacy, Us-China Relations, Cybersecurity Norms, Ethical Considerations.

"OMG you look amazing": A systematic examination of the text-based interactions surrounding UK adolescent girls' self-images on Instagram.

"OMG you look amazing": A systematic examination of the text-based interactions surrounding UK adolescent girls' self-images on Instagram.

A joint-training topic model for social media texts

The burgeoning significance of topic mining for social media text has intensified with the proliferation of social media platforms. Nevertheless, the brevity and discreteness of social media text pose significant challenges to conventional topic models, which often struggle to perform well on them. To address this, the paper establishes a more precise Position-Sensitive Word-Embedding Topic Model (PS-WETM) to adeptly capture intricate semantic and lexical relations within social media text. The model enriches the corpus and semantic relations based on word vector similarity, thereby yielding dense word vector representations. Furthermore, it proposes a position-sensitive word vector training model. The model meticulously distinguishes relations between the pivot word and context words positioned differently by assigning different weight matrices to context words in asymmetrical positions. Additionally, the model incorporates self-attention mechanism to globally capture dependencies between each element in the input word vectors, and calculates the contribution of each word to the topic matching performance. The experiment result highlights that the customized topic model outperforms existing short-text topic models, such as PTM, SPTM, DMM, GPU-DMM, GLTM and WETM. Hence, PS-WETM adeptly identifies diverse topics in social media text, demonstrating its outstanding performance in handling short texts with sparse words and discrete semantic relations.

Political Themes and Narratives in Videogame Cybertexts: Procedural Media Rhetoric in the Loop of Educational Practices

Introduction. General political education and a set of basic political science competences are important both for the formation of political culture of a particular person and for the balanced development of the state and strengthening of civil society. Video-game political simulators, acting as carriers and media-translators of multimodal political information, can become a legitimate participant of educational and political science discourse. Materials and Methods. The work is based on the theoretical basis of cybertextual-procedural interpretation of video games, funded by the concepts of E. Aarseth and I. Bogost. The multimodal-semiotic optics of video games is based on the synthesis of J. Havreljak and V. Toch's concepts. The teaching and learning (practical) part of the research is conceptually based on the Digital Game-Based Learning methodology of M. Prensky. Results. In the modus operandi of considering the language of video games and their media rhetorical techniques, the spectrum of dominant semiotic structures used to translate political content in video game cybertexts was explicated. The core of the political semiosphere consists of a bimodal sign structure consisting of iconic (visual-graphic and textual-letter) and procedural signs. The latter create a fundamentally new, interactive-action modality of information and communication interaction with a digital media text. The experimental part of the study confirmed the effectiveness of the DGBL methodology in studying political science topics and stories. The overwhelming majority of respondents (85%) noted the expediency of educational use of thematically profiled video game cybertexts. The final works demonstrated students' deep understanding of the dynamics and internal logic of political processes, as well as their connection with other spheres of social life. Qualitative testing showed a high (80%) mastery of thematically relevant issues and stories from the general political science course. At the same time, about 30% of students noted difficulties in mastering game interfaces and procedural mechanics of video games. KEYWORDS Conclusion. When integrating video-game cybertexts into the educational process, it is necessary to provide high-quality methodological and consulting support, as well as to have a technical (hardware and software) base and to take into account the factor of significant time expenditures required to master video games. Accordingly, it is recommended to use video-game political simulators within the framework of elective classes in political sciences. When using the DGBL methodology in the current political science teaching process, it is necessary to apply a modular approach, clearly limiting (chronologically and thematically) the work with video-game cybertexts.

Global icon or national star?: interpreting the contested star image of BLACKPINK’s Lisa in a transnational context

ABSTRACT This study examines how a transnational star’s persona is influenced and intervened by contesting nationalist and racist ideologies, with BLACKPINK’s Lisa serving as a case study. As an ethnic Thai K-pop star who has achieved global success, her identity can be framed either as a Thai star, a Korean pop star, or a global star. This overlap and potential conflict of multiple identities creates tensions surrounding her position among Korean and Thai audiences, which lead to questions pertaining to the articulation and negotiation between the local, national, and global. What contesting perspectives on nationality and ethnicity can we identify from Lisa’s star image in the context of transnational K-pop industry? What aspects of nationalism or racism can we discuss based on the production and consumption of her multi-faceted star image? To address these questions, this study investigates the construction of her star image in Thailand and Korea through an analysis of various media texts and discusses the nationalisation of transnationality and ethnic blindness within the context of transnational stardom.

Framing Analysis of News about Putri Ariani Winning the Golden Buzzer at America's Got Talent 2023 in Online Media Liputan6.com and Kompas.com

This research uses a qualitative descriptive method. Through a media text analysis approach with a focus on the issue of reporting on Putri Ariani. This type of qualitative research means that the researcher describes physical conditions and social conditions based on actual facts and data. Robert N Entman's analysis has four dimensions in the analysis, namely, Defining Problems, Diagnosing Causes, Moral Judgment/Evaluation, Treatment Recommendation. Where each news item will be analyzed one by one to find out the pattern of news construction created by the two media. Through media text analysis of Robert N Entman's framing model, the researcher draws conclusions regarding the formulation of the research problem. The author concludes that the frame built by Liputan6.com looks at the news. that Putri Ariani became the public spotlight after her stunning performance at America's Got Talent 2023. Meanwhile, the frame built by Kompas.com saw the news about Putri Ariani as a fact that could be seen in foreign media.

Goal framing of health promotion messages in contact lens wear.

Goal framing of health promotion messages in contact lens wear.

“And All I Gotta Do Is Act Naturally”

This article analyses musicians’ performances on film, using the iconic Beatles film Help! (Richard Lester 1965) as a foundational case study. The study investigates distinctions between film acting and musical performance while exploring the defining characteristics of musicians’ performances within the context of cinema. Building upon established concepts of persona (Auslander 2021), metafiction and metareference (Waugh 1984; Wolf 2009), the author introduces novel conceptual frameworks of “metaperformance” and “intramedial transmediality”. “Metaperformance” refers to the doubling of the act of performance, whereas “intramedial transmediality” describes the coexistence of diverse media texts within a single media text. Due to their off-screen musical persona, musical stars in films often provoke a pronounced transtextual and transmedial network and convey an implicit claim to reality. Compared to film actors, musical stars on film frequently provide metaperformances, embodying not only (fictional) characters but also performing their (“real”) musical persona within them. Additional case studies of Ed Sheeran’s performance in Danny Boyle’s Yesterday (2019) and the Spice Girls’ performance in Bob Spiers’s Spice World (1997) further clarify these theoretical insights. Working deductively, these films serve as contemporary illustrations of the theoretical concepts under examination. The findings of this essay contribute to an enriched understanding of the intricacies in musicians’ performances on film and shed light on the interplay between music, cinema, and artists’ on-screen personas.

Learnable Conjunction with Adjective Enhanced Model for Mandarin Sentiment Analysis of Social Media Text

The advent of research into neural networks has led to the widespread utilisation of deep learning methodologies in the domain of text sentiment analysis, owing to their formidable data processing and pattern recognition capabilities. In recent years, Transformer and its variant Bert have attracted considerable attention, and their performance has been demonstrated to be superior in practice. However, it is important to note that China's unique social media data set is characterised by a variety of languages, distinct cultural backgrounds, and intricate emotional expressions, which poses a significant challenge to the generalisability of models trained on more homogeneous data sets. Consequently, the conventional Transformer model's efficacy in feature extraction is constrained in certain scenarios, impeding the efficiency of its implementation. It is necessary to adapt and optimize the model further to account for the specific attributes of social media data in the task design. Nevertheless, within the framework of Chinese linguistics, the significance of word order and conjunction is inherently apparent. Conversely, the potential of adjectives in emotional expression is frequently disregarded. This study endeavours to investigate an innovative approach, underpinned by a Transformer encoder model, to construct a reinforcement model that can integrate and learn adjective features. The model will be evaluated through experimentation with three publicly available Chinese social media datasets. The experimental data demonstrate that the model can utilise the attention mechanism to not only identify the emotional tendency of keywords but also to effectively combine the positional characteristics of conjunctions and adjectives to obtain local details and contextual meanings of the text. This enhancement of the model's effectiveness in feature extraction is a significant contribution to the field.

EmoBERTa-X: Advanced Emotion Classifier with Multi-Head Attention and DES for Multilabel Emotion Classification

The rising prevalence of social media turns them into huge, rich repositories of human emotions. Understanding and categorizing human emotion from social media content is of fundamental importance for many reasons, such as improvement of user experience, monitoring of public sentiment, support for mental health, and enhancement of focused marketing strategies. However, social media text is often unstructured and ambiguous; hence, extracting meaningful emotional information is difficult. Thus, effective emotion classification needs advanced techniques. This article proposes a novel model, EmoBERTa-X, to enhance performance in multilabel emotion classification, particularly in informal and ambiguous social media texts. Attention mechanisms combined with ensemble learning, supported by preprocessing steps, help in avoiding issues such as class imbalance of the dataset, ambiguity in short texts, and the inherent complexities of multilabel classification. The experimental results on the GoEmotions dataset indicate that EmoBERTa-X has outperformed state-of-the-art models on fine-grained emotion-detection tasks in social media expressions with an accuracy increase of 4.32% over some popular approaches.

Analysis of the means of expressing doubt based on the Russian National Corpus

Summary Our aim is to use qualitative and quantitative analyses based on evaluation theory of news discourse from the Russian National Corpus to describe the underlying linguistic features of the means of expressing doubt and to identify some regularities that cannot be detected by intuitive analysis alone. Accordingly, in this article we introduce specialized marker and indirect marker, which were first used by Nikolskaya (2009: 144) in her own work. After analyzing some of the most representative dictionaries, we found that the use of indirect markers of doubt is also typical for expressing doubt (54 such units were found), namely: modal particles of doubt (вряд ли ‘hardly’, едва ли ‘barely’, вроде бы ‘seemingly’, etc.), parenthetical modal words (может быть ‘can be’, вероятно ‘probably’, возможно ‘maybe’, etc.), modal phrases, such as мне кажется ‘It seems to me that’, etc. We try to study their dynamic use and observe differences in frequency and methods of use in the texts of modern Russian-language mass media, and conduct their quantitative analysis.

The variation of the indefinite pronoun uno as a desubjectivizing resource by sex/gender

Abstract In Spanish, uno (‘one’) is third person pronoun whose referent is categorically a first-person singular referent of human nature: “Uno quiere salir adelante” (‘One wants to go onwards’). It represents the person speaking, but a second person or a nonspecified group of referents may also be included in the reference. Speakers refer to themselves by using this pronoun, but at the same time, they intend to give a universal reading to the content. The use of uno (‘one’) promotes a desubjectivizing style of discourse suited to achieve certain concrete communicative goals. Desubjectivization is a cognitive process based on the gradual cognitive salience or accessibility of personal referents. In this investigation, uno (‘one’) is analyzed as a desubjectivizing resource by observing its use across sex/gender factors. Sex/gender will be approached as a covariating factor with the use of the pronoun uno, considering that it is a grammatical choice that will contribute to the construction of a communicative style. Two pragmatic referential readings of uno were isolated and their analysis was performed in a corpus of social media texts (Facebook, X and Instagram). Results reveal that men tend to create a more desubjectivizing style according to their frequency of using unevenly both readings of uno in all and social media text. In turn, women showed an increasing percentage of just one of the uno’s reading on the social media X, thus, women tend to create a less desubjectivizing style than men.

Ideology, resistance, and sociopolitical dynamics in Indonesia: media narratives and resistance discourses on the chairman of the corruption eradication commission’s corruption case

Indonesia’s Corruption Eradication Commission (KPK) is an anti-graft organization established in 2003 that plays a vital role in preventing and eradicating corrupt practices in Indonesia. Ironically, KPK Chairman 2019–2023, Firli Bahuri, was named a corruption suspect on November 22, 2023. The incident of naming the KPK Chairman as a suspect by the police caused conflict between KPK and the Police Department and attracted the media’s attention. This article aims to discover how DetikCom, Kompas.com, and Tempo.co, produce texts and discourses of resistance by the KPK Chairman against being named a suspect. Researchers also tried to uncover the ideology and interests of the media in the production of corruption discourse in Indonesia based on the external context of media and politics in Indonesia. The author used the text analysis method with Norman Fairclough’s Critical Discourse Analysis to analyze 18 media texts. The researcher found that the resistance discourse packaged by Detikcom online media tends to support the law enforcement narrative by providing more space for the Police Department to convey evidence and processes carried out. Kompas.com emphasizes that Firli has the right to file a pretrial and that the legal process must be respected, giving the impression that the media respects Firli’s legal rights. Tempo.co emphasized the case’s legal aspects and processes, including Firli Bahuri’s pretrial submission. At the sociocultural level, there is a relationship between corruption cases in Indonesia, government policies, and public responses to the conflict discourse over the KPK chairman.

Strategies of Legitimisation and De-legitimisation in Polarised Representations of Climate Activists in Swedish Media

The number of media representations of climate activists has increased dramatically in Sweden recently. Some climate actions have attracted particular attention from Swedish media, such as when activists stopped traffic and disrupted public events. In this article, I present an analysis of media representations of both supporters and critics of climate activists in terms of strategies of legitimisation and de-legitimisation. The study is based on an examination of 98 Swedish media texts. The results reveal that risk-taking, self-sacrificing, heroisation, normalisation and personification are legitimisation strategies, while de-legitimisation strategies involve de-politicisation, rigid opposition, undermining the motive and stigmatisation. A conclusion is that the identified (de)legitimisation strategies are an expression of and contribute to increased polarisation between those who support the activists and those who criticise them. This polarisation takes different forms (partisan polarisation, affective polarisation and cultural polarisation) and risks contributing to an antagonistic debate where political opponents are transformed into enemies to be destroyed.

SafeSpeech: a three-module pipeline for hate intensity mitigation of social media texts in Indic languages

Warning: This paper contains some abusive text that might sometimes be found offensive.Identifying and mitigating hateful, abusive, offensive comments on social media is a crucial, paramount task. It’s challenging to entirely prevent such hateful content and impose rigorous censorship on social platforms while safeguarding free speech. Recent studies have focused on detecting hate speech, whereas mitigating the intensity of hate remains unexplored or somewhat complex. This paper introduces a cost-effective, straightforward, and novel three-module pipeline, SafeSpeech, for Hate Speech Classification, Hate Intensity Identification, and Hate Intensity Mitigation on social media texts. The initial module classifies text as either containing or not containing hate speech. Following this, the second module quantifies the intensity of hate associated with individual words within the classified hate speech. Lastly, the third module seeks to diminish the overall hatefulness conveyed in the text. A comprehensive experiment has been conducted using publicly available datasets in five Indic languages (Hindi, Marathi, Tamil, Telugu, and Bengali). The system undergoes thorough evaluation to assess its performance and analyze it in-depth using various automated metrics. We evaluated the performance of Hate Speech Classification using Precision, Recall, and F1 metrics. For Hate Intensity Identification, we use human evaluation. Recognizing the limitations of automated metrics in mitigating hate speech, we augment our experiments with human evaluation for Hate Intensity Mitigation, where three domain experts independently participated. BERTScore for final generated hate-mitigated texts and first classified hate texts across all languages consistently range between 0.96 and 0.99.

Quantitative research on aesthetic value of the world heritage karst based on UGC data: A case study of Huangguoshu Scenic Area.

The World Natural Heritage is a rare and irreplaceable natural landscape recognized by all mankind, with outstanding significance and universal value. Among them, the World Heritage Karst sites(WHKs) holds an important position due to its special natural beauty and aesthetic value. In the field of landscape evaluation, interdisciplinary and interdisciplinary cooperation using different methods has always been a research focus. However, there is still a gap in the evaluation of natural landscape aesthetic value based on UGC(User Generated Content) data and deep learning models. This article is based on a public perspective, using social media UGC data, crawling images and texts as data sources, and combining SegFormer deep learning models, ArcGIS spatial analysis, natural Language Processing Technology (NLP) and other methods to conduct quantitative research on aesthetic value. Research has found that: (1) Huangguoshu Scenic Area has an excellent natural environment, and landscape elements with high naturalness (vegetation, water) are more attractive to tourists, with diverse landscape combinations; (2) There is no complete positive correlation between tourist sentiment bias, landscape diversity, and vegetation coverage. Emphasis is placed on the aesthetic perception path from bottom to top, from the surface to the inside. The comprehensive emotional value is 14.35, and the emotional values are all positively distributed. The distribution density and extreme value of positive emotions are greater than those of negative emotions; (3) The emotional bias of tourists is directly related to visual sensitivity, showing a synchronous trend of change. The visual sensitivity of the Great Waterfall and Dishuitan areas is relatively high, mostly at I-II level sensitivity. This method enhances the data source channel, which is conducive to obtaining the correct tourist evaluation orientation. In traditional subjective landscape evaluation, rational parameter indicators are added to reduce the probability of error, provide data support for its natural beauty description, break through the time and space limitations of aesthetic evaluation, and provide scientific reference for quantifying the aesthetic value of other heritage sites.