Turkish News Research Articles

Examining Journalists’ Adoption of Social Media Tools in Contexts of Precarity

ABSTRACT Digital technologies have changed journalists’ work practices and their roles with far-reaching effects for news production. While many studies focus on the ways social media transforms editorial processes, much of this research examines fully democratic Western countries. We examine how journalists in Turkey, who occupy particularly precarious social positions, incorporate social media tools into their routines, and the effects of these tools on journalistic role conceptions. To understand this, we analyzed 10 interviews with Turkish editors and reporters from popular Turkish news outlets. These journalists say their roles are shifting from active news formation to passive information relay — they are becoming aggregators rather than reporters. And while social networking sites increase the volume of newsworthy ideas that journalists encounter, they also increase the pressure on editors and reporters to compete with other news outlets. The organizations we studied did not employ separate fact-checking teams or software for news verification, leading to workflow challenges around verifying large quantities of information collected from social media. Our findings suggest that journalists in precarious positions might be especially reliant on social media logics in their daily work since they lack the social, political, and occupational security to improvise or resist new practices and values.

Turkish Text Classification Based On Wrapper Feature Selection Using Particle Swarm Optimization

The vast majority of the digital era data is stored as text. Text mining is an integral part of data mining. Text classification (TC) is a natural language processing (NLP) operation often needed in text mining. This operation is needed in numerous kinds of research such as information retrieval, document classification, language detection, sentiment analysis, etc. According to the literature, the filter feature selection methods have often been applied to reduce the dimensionality of data in Turkish TC. However, the wrapper-based feature selection methods can provide better classification accuracies than the filter methods. Motivated by this idea, a Turkish TC method based on wrapper feature selection using particle swarm optimization algorithm (PSO) and multinomial naive bayes (MNB) classifier is proposed in this study. TTC-3600 Turkish news texts are used for TC in the experiments. The proposed method achieves a classification accuracy of 94.55% on TTC-3600 Turkish news text dataset by using stemming Tf-Idf features. Hence, it produces competitive accuracies to the cutting-edge Turkish TC methods.

Lead paragraphs of French and Turkish news items: Insights for French language learners

This study compared French and Turkish news items related to road accidents, specifically focusing on lead paragraphs. It aimed to provide insights for Turkish speakers learning French about reading comprehension and written expression at A2 and B1 levels. Data were collected from five newspapers in each country. French news items and Turkish news items showed similarities in word count but differed significantly in sentence and clause structures. French paragraphs tended to have longer single sentences with fewer clauses, while Turkish paragraphs had shorter sentences with more clauses. The study also explored the inclusion of informational components in lead paragraphs, such as accident scenes, types of vehicles, information about victims, circumstances of the accident, and casualty data. Results revealed variations between the two languages, emphasizing the importance of including information about casualties in lead paragraphs for effective communication in both languages.

Comparison of feature-based sentence ranking methods for extractive summarization of Turkish news texts

Comparison of feature-based sentence ranking methods for extractive summarization of Turkish news texts

Biological gender identification in Turkish news text using deep learning models

Biological gender identification in Turkish news text using deep learning models

Classification of News Texts from Different Languages with Machine Learning Algorithms

As a result of the developments in technology, the internet is accepted as one of the most important sources of information today. Although it is possible to access a large number of data in a short time thanks to the Internet, it is critical to analyze this data correctly. The need for text mining is increasing day by day by processing and analyzing the increasingly irregular text type data in the digital environment and classifying them in a meaningful way. In this study, news texts obtained from online German, Spanish, English and Turkish news sites were separated according to predetermined world, sports, economy and politics categories. The data set consisting of 4000 news texts was classified using 41 different machine learning algorithms in the Weka program. The highest successful classification was obtained with Naive Bayes Multinominal and Naive Bayes Multinominal Updateable algorithms, and 93.5% for German news texts, 93.3% for English news texts, 82.8% for Spanish news texts and 88.8% for Turkish news texts.

ON THE EFFECTIVENESS OF PARAGRAPH VECTOR MODELS IN DOCUMENT SIMILARITY ESTIMATION FOR TURKISH NEWS CATEGORIZATION

News categorization, which is a common application area of text classification, is the task of automatic annotation of news articles with predefined categories. In parallel with the rise of deep learning techniques in the field of machine learning, neural embedding models have been widely utilized to capture hidden relationships and similarities among textual representations of news articles. In this study, we approach the Turkish news categorization problem as an ad-hoc retrieval task and investigate the effectiveness of paragraph vector models to compute and utilize document-wise similarities of Turkish news articles. We propose an ensemble categorization approach that consists of three main stages, namely, document processing, paragraph vector learning, and document similarity estimation. Extensive experiments conducted on the TTC-3600 dataset reveal that the proposed system can reach up to 93.5% classification accuracy, which is a remarkable performance when compared to the baseline and state-of-the-art methods. Moreover, it is also shown that the Distributed Bag of Words version of Paragraph Vectors performs better than the Distributed Memory Model of Paragraph Vectors in terms of both accuracy and computational performance.

Analysis of Anti-Vaccination Discourse on Turkish Alternative News Websites

The media's information and enticement functions have grown increasingly important as a result of the Covid-19 pandemic. The purpose of this research on the anti-vaccination movement in Turkey was to better elucidate the different types of arguments that are deployed by key actors in the field of anti-vaccination, especially online. This research has analyzed the anti-vaccine movement, particularly in Turkey. More specifically, the argument that is deployed online by the alternative Turkish news websites is being analyzed. According to the results of the study, it was found that the news sites do not agree even when it comes to global issues such as pandemics. It is searched that alternative news websites have adopted the discourse of anti-vaccination without truly basing their arguments on scientific facts.

Turkish abstractive text document summarization using text to text transfer transformer

Text summarization is the process of reducing text size while preserving its key points. Thanks to this process, the reading time of the text is also reduced which contributes to reaching the desired information quickly, especially in today's world where time is much more important. In addition, summarization can be used to create a solution to extract outstanding information from the text. In this study, we focus on abstract summarization, which can draw more human like conclusions from the text. A summarization study was carried out on the data set that was collected from online Turkish news sources. Rouge and Bert-score performance metrics were used to compare the performance of this study using the text to text transfer transformer (T5) method. The precision values ​​of the Rouge-1, Rouge-2, Rouge-L and Bert-score performance metrics obtained in this study were found to be 0.6913, 0.6623, 0.7528 and 0.8718, respectively. Recall values ​​were 0.9210, 0.8917, 0.9183 and 0.9138, respectively. F measure values ​​were 0.7649, 0.7338, 0.8084 and 0.8913 respectively. Considering the success of the results obtained in the study, a method that can obtain successful results for Turkish text summarization is presented and the original dataset is made available to other researchers.

Türkçe Metinde Topluluk Öğrenme ve Genetik Algoritma Kombinasyonu Tabanlı Yazar Tahmini

The easiness of reaching information through the internet and social media and the expansiveness of opportunities for searching, copying, and spreading data have caused some problems in identifying an author for a specific text. A text carries the characteristic features of the person who wrote it, and these features can be used to identify its author. For this study, we are offering a method that is based on an approach using ensemble learning algorithm (ELA) and genetic algorithm (GA) for author identification in Tur-kish texts. The raw data set, which includes 40 authors and 3269 texts, was created from Turkish news websites and analyzed in pre-processing step. After, syntactic and structural analyses were done on the data and, in total, 6 different data sets were created. Each of the data sets was subjected to the feature selection process by using GA and ELA approach together. Each of the obtained data sets from the previous step was classified by using the ELA's bagging method which contains 5 different classifiers, namely, Naive Bayes, K-Nearest Neighbor, Artificial Neural Networks, Support Vector Machine, and Decision Tree. After applying the aforementioned processes to the raw data, the author identification approach reached 89% accuracy. The combination of ELA and GA has a strong potential to identify the author of a text.

Detecting Turkish Fake News Via Text Mining to Protect Brand Integrity

Fake news has been in our lives as part of the media for years. With the recent spread of digital news platforms, it affects not only traditional media but also online media as well. Therefore, while companies seek to increase their own brand awareness, they should also protect their brands against fake news spread on social networks and traditional media. This study discusses a solution that accurately classifies the Turkish news published online as real and fake. For this purpose, a machine learning model is trained with tagged news. Initially, the headlines were analyzed within the scope of this study that are collected from Turkish online sources. As a next step, in addition to the headlines of these news, news contexts are also used in the analysis. Analysis are done with unigrams and bigrams. The results show 95% success for the headlines and 80% for the texts for correctly classifying the fake Turkish news articles. This is the first study in the literature that introduces an ML model that can accurately identify fake news in Turkish language.

Author and genre identification of Turkish news texts using deep learning algorithms

Author and genre identification of Turkish news texts using deep learning algorithms

Establishing a Digital Belonging

Abstract A sense of belonging is one of the most important human needs. Immigrants might experience this differently because they tend to feel a lack of belonging to their host or home country. Nowadays, however, social media allows immigrants and diasporic groups to establish a digital sense of belonging. Members of the Turkish diaspora in Germany are inclined to use social media, especially Facebook, as a platform to connect with their roots, be informed about Turkish news, prove their Turkishness, and provide solidarity for each other. For this paper, we interviewed fifteen Turkish-Germans to get some insight into their perceptions about developing a digital belonging to Turkey through Facebook. We used discourse analysis to analyse their statements. The results suggest that Facebook allows Turkish-Germans to develop a digital belonging to Turkey through its specific features; however, it decreases their sense of belonging to Germany, as stated in the interviews.

Turkish press climate crisis coverage (2018–2019): elements of disconnect in discourses and the representation of solutions

AbstractRecent studies have shown that the media in developing countries recognizes the anthropogenic impact on climate change, while ignoring the mitigation and adaptation responsibilities of national political actors. This article addresses the discursive motives underlying the disconnect between the impact of climate change and the responsibility of those in positions of political power. This study analyzes Turkish news articles and columns on climate change published in three newspapers with different political orientations between June 2018 and January 2020, a period during which the school strike movement and other local uprisings and debates began. It claims that news related to national climate policy largely omits or obscures references to the anthropogenic causes of climate change, to the degree that the political responsibility of tackling it remains unaddressed and absolves readers and politicians from taking action. The article also aims to underline the impact of political parallelism in terms of the print media’s approach to the government’s neoliberal economic policy and its duty to tackle climate change. Finally, it argues that these approaches generate specific types of environmental discourses which are embodied in newspapers’ conceptions of nature, their solutions to climate change, and the actors of those solutions proposed by newspapers.

TwitterBulletin: An Intelligent and Real-Time Automated News Categorization Tool for Twitter

Social media platforms have become popular news sources thanks to their immense popularity and high speed of information dissemination. Using these platforms is essential for news organizations and journalists to track and discover news in digital journalism age. However, the abundance of meaningless data and the lack of organization on these platforms make it difficult to reach valuable news for journalists. In this paper, we create the first public dataset containing large number of real-world Turkish news tweets belonging to different news categories, to the best of our knowledge. We propose an artificial intelligence-based two-step approach to assist journalists for accessing the news shared by various sources on social media under the relevant categories like politics (elections, riots, etc.), health (pandemic, covid-19, etc.), etc. via a single platform by reducing the possibility of overlooking needed information. In the first step, we propose a machine learning based novel model for collecting and categorizing news posts on social media. We implement several traditional machine learning and deep learning based algorithms and evaluate their classification performance in terms of accuracy, precision, recall, and F1 score. In the second step, we develop a software tool, named TwitterBulletin, which automatically retrieves Turkish news tweets and groups them under news categories in real time by using the CNN classifier which achieves the best performance in the first step. The results show that the overall accuracy rate of TwitterBulletin is reasonably high and satisfactory despite the challenge of classifying short tweets.

Editorial

Dear Readers, It gives me great pleasure to announce the third regular issue of 2022. I would like to thank all the authors for their sound research and the editorial board for the extremely valuable reviews and suggestions for improvement. These contributions, together with the generous support of the consortium members, enable us to run our journal and maintain its quality. Still, I would like to expand our editorial board: If you are a tenured associate professor or above with a good publication record, please apply to join our editorial board. We are also interested in receiving high-quality proposals for special issues on new topics and emerging trends. In this regular issue, I am very pleased to introduce four accepted papers involving nine authors from four different countries. Sedef Demirci and Seref Sagiroglu from Turkey introduce TwitterBulletin which is a real-time automated news categorization tool for Twitter based on a novel machine learning approach built on a public dataset with a large number of real-world Turkish news tweets. Also in the social media domain, Vitor Garcia dos Santos and Ivandré Paraboni from Brazil apply pre-trained language models to the classification of personalities from texts, focusing specifically on the Myers-Briggs personality model. Mohd Khaled Y. Shambour and Esam A. Khan from the Kingdom of Saudi Arabia discuss in their paper a hyper-heuristic approach to optimize the distribution process of pilgrims over Mina tent camps by selecting one among four predefined low-level heuristics to generate a new one, and discuss solution quality and convergence rate. Nahia Ugarte, Mikel Larrañaga and Ana Arruarte from Spain conduct a systematic mapping review that explores the use of recommender systems in formal learning stages. Enjoy Reading!

The (re)presentation of interpreters in the Turkish media: a diachronic and cross-setting analysis

ABSTRACT This study explores the representation of interpreters in the Turkish news media by analysing the online editions of the three most circulated Turkish newspapers between June 2017 and March 2021. It looks at what triggers a media discourse on interpreters and how interpreters are represented in the news media. It also juxtaposes the current portrayal of conference and diplomatic interpreters with the media coverage of conference interpreters in two previous studies. The results indicate that although references to interpreters working in legal settings are by far the most numerous, these references are almost always in passing, giving interpreters in these settings no visibility or voice. The same applies to interpreters working in refugee settings. In contrast, news items on sign language and football interpreters accord significant visibility and voice to these interpreters. News reports on conference and diplomatic interpreters, on the other hand, seem to reflect a marked shift towards negative coverage, characterised by the absence of interpreters’ voices and a media discourse that contests their neutrality, the latter view of interpreters being at least partly related to political tensions in the country.

Improving automated Turkish text classification with learning‐based algorithms

AbstractText classification is the process of determining categories or tags of a document depending on its content. Although text classification is a well‐known process, it has many steps that require tuning to improve mathematical models. This article provides a novel methodology and expresses key points to improve text classification performance using learning‐based algorithms and techniques. First, to check the effectiveness of the proposed methodology, we selected two public Turkish news benchmarking datasets. Then, we performed extensive testing using both supervised machine learning algorithms and state‐of‐art pre‐trained language models. The experimental results show that our methodology outperforms previous news classification studies on these benchmarking datasets improving categorization results based on F1‐score. Therefore, we conclude that the presented methodology efficiently improves the classification results and selects the feasible classifier for a given dataset.

A comparative study of neural machine translation models for Turkish language

Machine translation (MT) is an important challenge in the fields of Computational Linguistics. In this study, we conducted neural machine translation (NMT) experiments on two different architectures. First, Sequence to Sequence (Seq2Seq) architecture along with a variation that utilizes attention mechanism is performed on translation task. Second, an architecture that is fully based on the self-attention mechanism, namely Transformer, is employed to perform a comprehensive comparison. Besides, the contribution of employing Byte Pair Encoding (BPE) and Gumbel Softmax distributions are examined for both architectures. The experiments are conducted on two different datasets: TED Talks that is one of the popular benchmark datasets for NMT especially among morphologically rich languages like Turkish and WMT18 News dataset that is provided by The Third Conference on Machine Translation (WMT) for shared tasks on various aspects of machine translation. The evaluation of Turkish-to-English translations’ results demonstrate that the Transformer model with combination of BPE and Gumbel Softmax achieved 22.4 BLEU score on TED Talks and 38.7 BLUE score on WMT18 News dataset. The empirical results support that using Gumbel Softmax distribution improves the quality of translations for both architectures.

Discussions on Freedom of the Press and Freedom of Expression in Türkiye: Analyzing the News Coverage of the “Disinformation Bill”

This study aims to examine the news coverage of the draft “disinformation bill” to reveal how the media as a key actor deal with the media freedom matter in the framework of free speech and focuses on the importance of press freedom as a fundamental principle in a democratic society. Analyzing the news coverage of the new bill provides insight into discussing how the divided media landscape and media ownership interpret the principle of press freedom. This study argued that critical news coverage of the legislative regulations on media freedom and combating disinformation can contribute to a more democratic public sphere and discussion. It needs more discussion on what should be done to prevent disinformation. This study concluded that the critical news coverage in regard to the disinformation bill was more included Cumhuriyet and Birgün which are anti-government and independent left-wing media defended the media freedom notion more than Habertürk and Sabah which are mainstream and pro-government. This result is connected to the polarization of Turkish news media.