News Corpus Research Articles

Ontological knowledge inferring approach: Introducing Directed Collocations (DC) and Joined Directed Collocations (JDC)

The growing need of utilizing unstructured knowledge embedded in open-domain natural language text into machine-processable forms requires the induction of hardly extracted structured knowledge into knowledge bases which makes the Semantic Web vision a reality. In this context, ontologies, and ontological knowledge (triples) plays a vital role. This research introduces two novel concepts named Directed Collocation (DC) and Joined Directed Collocation (JDC) along with a methodical application of them to infer new ontological knowledge. Introduced Quality-Threshold-Value (QTV) parameter improves the quality of the inferred ontological knowledge. Having set a moderate value (3) for QTV, this approach inferred 95,491 new ontological knowledge from 43,100 triples of open domain Sri Lankan English news corpus. Indeed, the outcome was approximately doubled in size as the source corpus. Some inferred ontological knowledge was identical with the original corpus content, which evidences the accuracy of this approach. The remaining were validated using inter-rater agreement method (high reliability) and out of which around 56% were estimated as effective. The inferred outcome which is in the triple format may use in any knowledge base. The proposed approach is domain independent. Thus, helps to construct/extend ontologies for any domain with the help of less or no human specialists.

News sharing as a measure of media alignment

In this note we introduce a new approach to measure media alignment derived from the story-sharing behavior of journalists. We use a large corpus of online news stories from two leading Hungarian news sites and estimate alignment scores for a large number of outlets that they cite. To the extent that journalists are more likely to cite ideologically proximate sources, our measure can be used to compare a large number of media outlets on a political — in our case government vs. independent — space. We demonstrate the use of this approach with two empirical applications. First, we show that our alignment scores successfully capture known ideological variation across outlets at a single point in time. Second, we demonstrate that quarterly estimates of alignment for a captured outlet change dramatically following an abrupt change in ownership.

A Study of Chinese News Headline Classification Based on Keyword Feature Expansion

Existing work generally classifies news headlines as a matter of short text classification. However, due to the strong domain nature and limited text length of news headlines, their classification results are usually determined by several specific keywords, which makes the traditional short text classification method ineffective. In this paper, we propose a new method to identify keywords in news headlines and expand their features from sentence level and word level respectively, and finally use convolutional neural networks (CNN) to extract and classify their features. The proposed model was tested on the Sogou News Corpus dataset and achieved 93.42% accuracy.

On the Margins: Exploring Minority News Media Representations of Women during the COVID-19 Pandemic

The COVID-19 pandemic has had a disproportionately negative affect on women, especially women from minoritized groups. Minority news media are an important information source for these groups, when it comes to providing alternative views, and health related information. Whilst the poor representation of women in COVID-19 related mainstream news coverage is acknowledged, little is known regarding the representation of women in digital minority news content, during the pandemic and beyond. Considering this gap, we examine how women have been represented within a diverse corpus of minority news, and explore how these representations serve to bridge between different social groups. Critically analyzing the representation of these marginalized groups offers a lens through which we can better understand the function of minority news media in a democracy. Using critical discourse analysis, this study examines three online minority newspapers and one podcast from the US. Findings show that women are covered in an inclusive and empowering way, containing perspectives and concerns unique to the minority group, strengthening identities, platforming community-specific issues, communicating a call to action, and promoting intersectional solidarity. These representations also reveal the complex tensions between counterhegemonic and dominant publics which minority news media sources constantly negotiate through their content.

Использование психологических терминов в различных типах дискурсов (на примере новостного корпуса)

В работе рассматривается употребление психологических терминов в непрофессиональных дискурсах с позиции типа интеракции, задействованных лингвистических механизмов и передаваемой оценки. Актуальность исследования обусловлена активным проникновением терминов психологии в различные типы дискурса: лексика, которая употреблялась изначально исключительно профессионалами при общении на профильные темы, прочно укрепилась в лексиконе людей, которые не связаны с психологией. Результаты исследования психологических терминов и механизмов их переносов в непрофессиональные дискурсы, проведенного на примере корпуса новостных статей, выявили, что: во-первых, самыми частотными являются термины с отрицательной оценкой, во-вторых, наблюдается отчетливая тенденция к переносу терминов в непрофессиональные дискурсы на основе метафоры с сохранением оценки. The work deals with the psychological terms in non-professional discourses with regard to the type of interaction, linguistic devices engaged and evaluation expressed. The relevance of the research is caused by the active penetration of psychological terminology in different types of discourses. The vocabulary initially used exclusively by professionals while communicating on fieldspecific topics has firmly entered the lexicon of individuals not engaged in psychology. The findings from the research of the psychological terms and mechanisms for transferring them to nonprofessional discourses performed by the material of a news article corpus revealed that: firstly, the most frequent are terms with the negative evaluation; secondly, there is a clear tendency for metaphor-based transfer of the terms into non-professional discourses keeping evaluative components.

Koronavirustaudin kehysmalli kriisiviestinnässä

During the surge of the coronavirus disease COVID-19, the need to disseminate accurate information and bridge the communicative gap between professionals and laypeople was urgent. During a pandemic, the role of mass media as well as translators and interpreters is essential in the distribution of information to all citizens. To facilitate the work of translators and interpreters, the Terminology Coordination Unit of the European Parliament has published the COVID-19 event frame (Termcoord 2020), which is based on Frame-based Terminology (see e.g. Faber 2015). In this paper, I report the results of the study on the creation of a parallel COVID-19 event frame based on a Finnish online corpus of news (Yle) and a Finnish ad hoc corpus of newspaper articles (HS). The study employed the methods of Frame-based Terminology. It is our aim that the frame could be useful in crisis communication for different stakeholders, including translators and interpreters.

Engagement Features in Business Newspapers Corpus: A Comparative Study of Business News and Business Articles

The study explores the Pakistani business newspaper corpus for the presence of engagement features—one of the systems of interactional metadiscourse. Business corpus of 1 million words comprised Pakistani business news and business articles collected from the four well-reputed Pakistani English newspapers, i.e., The Daily Times, The Nation, The Business Recorder, and The Dawn. It was found that the writer presents his ideas and arguments and engages his/her reader by employing different engagement features Furthermore, in the corpus of business articles, the writers were found keener to be informative and engaging than they were for the corpus of business news. The results of the study can potentially help future researchers in understanding the use of Engagement markers in other business corpus texts. In addition, educationists and syllabus designers may use the findings of the study for developing better teaching materials.

Tracing Linguistic Evidence of Cultural Diversity: A Corpus-based Study

Various factors, including culture, indigenous languages, and social norms, shape English being spoken and written in a foreign context. To trace the influence of these various factors, several registers/ genres can be investigated. However, the newspaper register is the most suitable as it is close to everyday events to study how the language has undergone a change. This study seeks to investigate the influence of these factors on the English used in a foreign context and explores the patterns of linguistic variation in the corpus of Cultural news from South Asian countries. A corpus from South Asian countries was compiled. Biber's (1988) multidimensional model was used. ANOVA was applied to find the statistically significant differences. The results indicate that Cultural press reportage of all the countries varies in producing explicit, non-argumentative and abstract discourse. The results indicate that as the cultural press reportage of every country is different, English used in a foreign context differs in producing newspaper discourse.

Investigating positive/negative bias in Canadian newspapers through translation: a study of ‘confidence’ in a corpus of business news

ABSTRACT Used as gauges of positivity and negativity in the press, the words ‘confidence’ and ‘confiance’ are analysed in a 2.7-million-word bilingual sub-corpus of CAPCOF, a 9-million-word corpus of economic and financial news extracted from seven Canadian newspapers. Focused on the year 2008 of the financial crisis, results show that the English-language media foresee a better tomorrow and are more positive than their French-language counterparts. Ultimately, however, a positivity bias emerges in both languages. The critical discourse analysis shows that the main point of convergence between the two linguistic media communities is the interdiscursive dynamics through which the informative discourse of the media and the promotional discourse of the financial establishment meld together seamlessly. In this study, translation is employed in its broadened definition, encompassing intralingual activity as the locus of interdiscursivity, rewording and recontextualization.

The Middle-class Card: A Corpus-based Study of the Characteristics of the Middle Class in US News

The ups and downs of the American middle class led several social scientists to rethink and reflect on their conditions, yet the characteristics of how the middle class are represented in the news have been hardly examined from a linguistic perspective. This study used a synergy of tools of Corpus Linguistics and Critical Discourse Analysis, guided by Norman Fairclough’s 3-D model (1989), to investigate the collocation profile of the middle class and the ideology underpinning its representation in US web news discourse from 2010 until 2022. The study relied on a specialized news corpus that had 1,253,678 tokens, including middle class (30,975 times) and middle-class (23,587 times). The findings showed that the US web news was consistent in constructing the characteristics of the middle class during the past twelve years. Under neoliberal policies, the idea of classism is an intrinsic dividing system in American society, and the notion that the middle-class family is an economic unit seems to be ubiquitous. Also, the middle class are depicted as the economic unit that establishes stability and the political card that politicians use in their agendas to win the majority of votes. In spite of the political and economic significance of the middle class, they are mostly passivated by minimizing their human agency and downplaying their roles as doers of social actions. The study can be a part of a branch of Applied Linguistics that focuses on the relationship between the science of economics and the science of language.

Creating and analysing a multimodal corpus of news texts with Google Cloud Vision's automatic image tagger

This study describes the creation and analysis of a small multimodal corpus of British news articles about obesity, where tags were assigned to images in the articles using the automatic tagger Google Cloud Vision. In order to illustrate the potential for analysis of image tags, the corpus analysis tool WordSmith was used to identify differences between newspapers in the ways that obesity was framed. Three forms of analysis were carried out – the first simply compared keywords across the newspapers, the second examined key visual tags and their collocates associated with each newspaper, while the third incorporated a combined analysis of words and image tags. The three analyses produced complementary findings, indicating the value in using Google Cloud Vision in creating and analysing multimodal corpora. The paper ends by reflecting on the method undertaken, while considering how additional research could improve our understanding of image tagging.

Improving Domain-Generalized Few-Shot Text Classification with Multi-Level Distributional Signatures

Domain-generalized few-shot text classification (DG-FSTC) is a new setting for few-shot text classification (FSTC). In DG-FSTC, the model is meta-trained on a multi-domain dataset, and meta-tested on unseen datasets with different domains. However, previous methods mostly construct semantic representations by learning from words directly, which is limited in domain adaptability. In this study, we enhance the domain adaptability of the model by utilizing the distributional signatures of texts that indicate domain-related features in specific domains. We propose a Multi-level Distributional Signatures based model, namely MultiDS. Firstly, inspired by pretrained language models, we compute distributional signatures from an extra large news corpus, and we denote these as domain-agnostic features. Then we calculate the distributional signatures from texts in the same domain and texts from the same class, respectively. These two kinds of information are regarded as domain-specific and class-specific features, respectively. After that, we fuse and translate these three distributional signatures into word-level attention values, which enables the model to capture informative features as domain changes. In addition, we utilize domain-specific distributional signatures for the calibration of feature representations in specific domains. The calibration vectors produced by the domain-specific distributional signatures and word embeddings help the model adapt to various domains. Extensive experiments are performed on four benchmarks. The results demonstrate that our proposed method beats the state-of-the-art method with an average improvement of 1.41% on four datasets. Compared with five competitive baselines, our method achieves the best average performance. The ablation studies prove the effectiveness of each proposed module.

Keyword extraction from news corpus by deep learning in the context of internet of things

Keyword extraction from news corpus by deep learning in the context of internet of things

VARIASI DAN POLA KOLOKASI VERBA DALAM MAKNA ‘MEMBERIKAN INFORMASI’ PADA KORPUS BERITA

There are many variations of verb in the meaning of 'providing information’ in the news text. This diversity of verbs in the same field of meaning often creates similar meanings from one to another. In sentences, these variation of verbs not only functioned to enrich the vocabulary and the beauty of dictions but also can show its function in sentences. It’s relevant with Chaer's statement (2012) which said that utterances or words of synonymous are never have the same meaning. It means, each word has its role and function in giving meaning to a sentence. This research was conducted to look further at the relationship of meaning between verbs which are in the same meaning field. The meaning relations which observed are the similarities and differences between these verbs by looking at the collocate that comes after those verbs. The collocate that accompanies the verb can give an idea of ​​the verb pattern in a sentence. This research using a corpus-based application with the theme 'Reconstruction of the Murder of Brigadier J by Ferdy Sambo'. The final results show that out of thirty published reports, fourteen variations of the verb which means 'providing information' with different collocate patterns are found. This research reinforces Chaer's (2012) statement that some verbs which have a close meaning also have their pattern even though they are in the same meaning field. Keywords: news corpus, verb variations, collocations.

Hindi to English Multimodal Machine Translation on News Dataset in Low Resource Setting

This work proposes a multimodal Hindi-to-English machine translation on a news corpus by integrating multiple input modalities. The experimental dataset comprises of an image from a news article and its caption in the English-Hindi language. The parallel English-Hindi dataset is constructed by translating the image caption from English-language news articles into Hindi. Unlike standard images used for NLP tasks, images in news articles frequently depict significant events, locations, or individuals. The caption provides a more extensive description of the image, resulting in the inclusion of multiple identified entities. Our methodology consists of encoders for each input modality, allowing for simultaneous consideration of both text and image. We undertake two tests to examine how effectively the image content improves MT systems when 1) the entire image is considered first and 2) the largest object recognized in the image is considered. We employ Byte Pair Encoding (BPE) to represent the text, and the visual attributes are extracted from the image with VGG-19, a pre-trained CNN model. In terms of BLEU and chrF, our Multimodal Machine Translation (MMT) systems are superior than the baseline unimodal NMT system. Our MMT system outperforms the NMT model by +1.8 BLEU and +0.03 chrF.

Mapping Climate Themes From 2008-2021—An Analysis of Business News Using Topic Models

India and other developing economies are receiving more attention in the context of climate change due to their rapid rates of economic expansion and large populations. In terms of absolute emissions, India surpassed China and the U.S. in 2018 to become the third-largest emitter. Solving this wicked problem calls for climate action across the stakeholder spectrum involving governments, business communities, and citizens. While extant literature has focused significantly on the role of governments and individual perceptions, the business sector needs to be more represented. In this study, we consider business news media as a platform that reflects the industry engagement in climate change and as a source of information on climate change for business decision-makers. Hence, understanding the topic and themes in the nexus of climate and business is important to evaluate the business sector’s stance towards climate change and how it has evolved. This work explores business news related to climate change using natural language techniques. We first experiment with three topic-modeling techniques, such as LDA, NMF, and BERTopic, on the business news and two more benchmark news datasets. Our test data is derived from digital news archives of ’The Economic Times – India’s leading business news daily. We evaluate the performance based on quantitative metrics commonly used for topic models. We choose the algorithm that provides the highest precision for climate-specific information represented by the test dataset. We then apply the algorithm with the best performance, as evaluated by the experiment, to a large corpus of Indian climate news from E.T. spanning from 2008 - 2021. We present how different themes, including industry engagement, evolved over the last two decades. The results suggest that climate cooperation has the highest contribution in the corpus, with other themes on resource management, energy and business gaining traction in recent years.

Ontology-Based News Linking for Semantic Temporal Queries

Daily newspapers publish a tremendous amount of information disseminated through the Internet. Freely available and easily accessible large online repositories are not indexed and are in an un-processable format. The major hindrance in developing and evaluating existing/new monolingual text in an image is that it is not linked and indexed. There is no method to reuse the online news images because of the unavailability of standardized benchmark corpora, especially for South Asian languages. The corpus is a vital resource for developing and evaluating text in an image to reuse local news systems in general and specifically for the Urdu language. Lack of indexing, primarily semantic indexing of the daily news items, makes news items impracticable for any querying. Moreover, the most straightforward search facility does not support these unindexed news resources. Our study addresses this gap by associating and marking the newspaper images with one of the widely spoken but under-resourced languages, i.e., Urdu. The present work proposed a method to build a benchmark corpus of news in image form by introducing a web crawler. The corpus is then semantically linked and annotated with daily news items. Two techniques are proposed for image annotation, free annotation and fixed cross examination annotation. The second technique got higher accuracy. Build news ontology in protégé using Ontology Web Language (OWL) language and indexed the annotations under it. The application is also built and linked with protégé so that the readers and journalists have an interface to query the news items directly. Similarly, news items linked together will provide complete coverage and bring together different opinions at a single location for readers to do the analysis themselves.

PROMOTION AND DISSEMINATION OF WULIANGYE YIBIN CULTURE IN THE ENGLISH-SPEAKING WORLD BASED ON PYTHON SEMANTIC ANALYSIS

The promotion and dissemination of Chinese Baijiu globally has become a hot topic in recent years. As Sichuan Baijiu gains popularity in both the international and domestic markets, the exploration of its global road is becoming increasingly important. This study focuses on the global path of Wuliangye Yibin culture and aims to propose strategies for promoting and disseminating it in the English-speaking world. This study makes use of an online English corpus and Python natural language processing technology to create a small Wuliangye Yibin English news corpus (WNC). This study then analyzes the frequency distribution and emotional tendency of adjectives related to the corpus, ultimately promoting and disseminating Wuliangye Yibin culture in Englishspeaking world. These include enhancing Wuliangye Yibin’s online presence, improving the quality of promotional materials translations, and targeting specific English-speaking audiences. Overall, this research serves as a valuable refence for the promotion of Chinese Baijiu culture on the global stage while adhering to China’s laws and regulation

Discriminating AI-generated Fake News

Generating fake news has become a prevalent issue in the digital age, with profound consequences on society, politics, and media. Due to an exponential increase in the use of unverified information, spread on social media by users with different backgrounds, a significant task facing natural language processing is real-time recognition of generated fake news. The objective of this paper is to investigate the feasibility of utilizing artificial intelligence (AI) for the automatic generation of fake news that possesses a sufficiently high level of quality to serve as a synthetic corpus. These automatically generated texts are subsequently used to improve fake news detectors, evaluated on different publicly available datasets fake news datasets, but also against a novel fake news corpus, especially developed for this research.

Semantik Prosodi Nomina Indonesia dalam Korpus Wacana Berita Arab Internasional Oman 2013

<p class="BodyA"><strong>The prosodic semantic analysis in this research focuses on the aura of the lexical and contextual meaning of words that collocate with the noun 'Indonesia', in the international news corpus of Omani Arabic "KALIMAT a Multipurpose Arabic Corpus, 2013". Using the AntConc corpus, two kinds of "Indonesian" nouns were found, "Indonesia" in the indefinite form /indu:ni: niyyah/ and in the definitive form /al-indu:ni: niyyah/. However, this study focuses on the noun "Indonesia" in the definitive form /al-indu:ni: niyyah/. The theory that is used in this study is the prosodic semantic theory proposed by Louw (1993), which states that the coloring or aura of the meaning of a form, is imbued with its collocations. The research design is descriptive qualitative and quantitative which is used to maximize the semantic analysis of the prosody of the research object. Based on the results of corpus data processing, it was found that 8 (eight) nouns collocated with 'Indonesian' nouns, namely /ba: li:/, /jazz: rah/, /asy-syurthah/, /al-Jindra:l/, /al -wathaniyyah/, /al-Musalaha/, /al-qua:t/, and /al- Hukumah/. From the results of the analysis, the authors found that the collocation of the noun 'Indonesia' is not only a single noun, it is also a phrase that gives a neutral and negative aura to the noun 'Indonesia'.</strong></p><p class="BodyA"><strong><em>Keywords</em></strong><strong> - </strong><em>Prosody Semantics, Arabic Corpus, Collocation, Indonesia</em></p>