Content Censorship Research Articles

Linguistic Features of Multilingual Hate Speech in the Online “Bardagulan”

"Bardagulan" is the slang used in the Philippines to describe an online dispute wherein Filipino netizens trade hate speech. The term was derived from the Tagalog word "Bardagul" which means conflict and is now employed to denote differing viewpoints (Yaneza, 2022). During the 2022 Presidential Election, a variety of hate speech was heard and seen, prompting the researcher to determine the linguistic characteristics of hate speech on Philippine social media. To demonstrate the complexity of hate speech in the Philippines, the researcher devised a research question centered on various linguistic features such as form, type, social aggression, impoliteness strategies, and the structure of hate speech to improve the understanding of how hate speech functions in actual online social discourse. It is crucial that all social media users possess knowledge regarding the linguistic features that distinguish hate speech, so that they may effectively report unpleasant remarks posted online. Multimodality and content analysis were also utilized to analyze 190 specimens of hate speech since the researcher considers hate speech as a multimodal phenomenon. This study suggests that the distinctive linguistic characteristics of multilingual "Bardagulan" hate speech contribute to the challenges encountered in eradicating hatred within the realm of Philippine social media. It was found that social media users in the Philippines employ grammatical and linguistic innovations to conceal profanity and circumvent digital content censorship. Moreover, the primary results indicate that behavior of Filipino internet users displayed via impoliteness in words, images, and emojis promote the proliferation of hate speech since they have a propensity to engage in confrontational discourse, with an aim to persuade the audience and to defend their own points of view.

Development and Regulatory Trends of China’s Online Live Streaming Industry

This article provides a comprehensive examination of the regulatory framework and government efforts in China’s online live streaming industry. It begins by outlining the relevant regulations and policies, encompassing content censorship, licensing requirements, user protection, and advertising regulations. The article then delves into the specific measures taken by the government to regulate content and safeguard user interests, including enhanced content censorship, implementation of licensing and registration systems, reinforcement of cybersecurity and data protection, as well as education and guidance initiatives. Emphasizing the significance of compliance with these regulations for the sustainable development of the industry, the article encourages platforms and content creators to actively cooperate with regulatory authorities and adhere to the provided guidelines, thereby fostering a healthy and thriving online live streaming ecosystem in China.

How to Find Orchestrated Trolls? A Case Study on Identifying Polarized Twitter Echo Chambers

Background: This study presents a graph-based, macro-scale, polarity-based, echo chamber detection approach for Twitter. Echo chambers are a concern as they can spread misinformation, and reinforce harmful stereotypes and biases in social networks. Methods: This study recorded the German-language Twitter stream over two months, recording about 6.7M accounts and their 75.5M interactions (33M retweets). This study focuses on retweet interaction patterns in the German-speaking Twitter stream and found that the greedy modularity maximization and HITS metric are the most effective methods for identifying echo chambers. Results: The purely structural detection approach identified an echo chamber (red community, 66K accounts) focused on a few topics with a triad of anti-Covid, right-wing populism and pro-Russian positions (very likely reinforced by Kremlin-orchestrated troll accounts). In contrast, a blue community (113K accounts) was much more heterogeneous and showed “normal” communication interaction patterns. Conclusions: The study highlights the effects of echo chambers as they can make political discourse dysfunctional and foster polarization in open societies. The presented results contribute to identifying problematic interaction patterns in social networks often involved in the spread of disinformation by problematic actors. It is important to note that not the content but only the interaction patterns would be used as a decision criterion, thus avoiding problematic content censorship.

System of governmental control over the information space in the U.S.

After the 2016 election disinformation has been securitized in the U.S. This precedent, which intertwined disinformation with the foreign influence in public opinion, laid a foundation for the emerging system of governmental control over the information space in the U.S. This article analyzes American official institutes charged with countering disinformation and foreign influence as well as the ways in which such activities are reinforced by input from tech companies and leading social media. Author studies how the limits of acceptable content moderation extended over the years. The first stage included public and private efforts to moderate content related to elections and coming from the foreign sources. During the second stage the content censorship started to be applied to topics other than the elections but nevertheless coming from abroad. The third stage – the current one – may be called a period of limitless control since the respective action has been taken by the government and social media both towards all kinds of information coming from foreign and domestic sources. Article concludes that the U.S. is entering the new electoral period amidst the existence of unprecedentedly obvious, well-established and effectively functioning system of governmental control over the internal information space.

Building a Personalized Model for Social Media Textual Content Censorship

Social media users often suffer from the problem of content over-disclosure. Most existing studies attempt to solve this problem by recommending proper audiences for users when sharing content. However, the audience management strategy cannot filter out sensitive information from the post and narrow the scope of content permeation. On the contrary, this paper conducts research from the content perspective and aims to design a content censorship model to help users evaluate the publicity of a post and find the sensitive information from it. The user can revise the content accordingly to achieve goals of sensitive information protection and broader content permeation. For this intention, we first built a dataset to explore the factors related to the public level of a post and the sensitive information. Based on the findings, a novel personalized multi-task content censorship model was built using several state-of-the-art deep learning techniques such as Seq2Seq and Co-training. We also implemented a prototype, i.e. a Browser plugin-based content censorship tool, by utilizing Weibo as a research site. Our model and its prototype were evaluated through automatic and human evaluations. The automatic evaluation suggests that our model outperforms the baseline methods on several metrics including precision, recall, and F1-score. The human evaluation also reveals that our model and prototype play an important role in helping users identify sensitive information. Based on these results, we proposed several insights for the future design of the social media content censorship system.

Bad actors never sleep: content manipulation on Reddit

ABSTRACT The self-proclaimed front page of the internet, Reddit is a many faceted site, described variously as an assemblage of online communities, a social news aggregation website and a social media site. Reddit is one of the most popular sites online both in terms of visitation and interaction. Central to Reddit’s popularity is its active community of users who contribute posts, comments and votes to subreddits resulting in a series of constantly changing bulletin boards comprised of over 3 billion submissions annually. Since its inception in 2005, Reddit has generated polarized engagement and responses to this community activity from wildly positive to highly toxic. Recent developments around perceived political manipulation of and on Reddit and accusations of content manipulation and censorship have raised questions of trust in the Reddit platform. Since 2019 Reddit has sought to identify and mitigate bad actor behaviour, address content manipulation and maintain the site’s authenticity and the historical context of Reddit’s attempts to address each is of central concern to this article. How this history then intersects with contemporary ideas of digital resignation and commons-based models of management informs analysis in the article on moderation practices, opens up new possibilities for future research into concepts of platform literacy and structural (digital) imagination.

A Study on Extent of Awareness Among College Students in Security and Privacy Issues in Social Media

Today Social Media is an integral part of many people’s lives. Most of us are users of one or many of these such as Facebook, Twitter, Instagram, LinkedIn etc. Social media networks are the most common platform to communicate with our friends, family and share thoughts, photos, videos and lots of other information in the common area of interest. Privacy has become an important concern in social networking sites. Users are not aware of the privacy risks involved on social media sites and they share their sensitive information on social network sites. While these platforms are free and offer unrestricted access to their services, they puzzle the users with many issues such as privacy, security, data harvesting, content censorship, leaking personal information etc. This paper aims at analyzing, the major users of social media networks, namely, the college students. It was intended to assess the extent the consumers’ are aware of the risks of free usage and how to mitigate against these privacy issues.

Design and Implementation of Fast Spoken Foul Language Recognition with Different End-to-End Deep Neural Network Architectures.

Given the excessive foul language identified in audio and video files and the detrimental consequences to an individual’s character and behaviour, content censorship is crucial to filter profanities from young viewers with higher exposure to uncensored content. Although manual detection and censorship were implemented, the methods proved tedious. Inevitably, misidentifications involving foul language owing to human weariness and the low performance in human visual systems concerning long screening time occurred. As such, this paper proposed an intelligent system for foul language censorship through a mechanized and strong detection method using advanced deep Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) through Long Short-Term Memory (LSTM) cells. Data on foul language were collected, annotated, augmented, and analysed for the development and evaluation of both CNN and RNN configurations. Hence, the results indicated the feasibility of the suggested systems by reporting a high volume of curse word identifications with only 2.53% to 5.92% of False Negative Rate (FNR). The proposed system outperformed state-of-the-art pre-trained neural networks on the novel foul language dataset and proved to reduce the computational cost with minimal trainable parameters.

Privacy risks with smartphone technologies when using the mobile Internet

This paper presents the results of a study about privacy risks when communicating and using the mobile Internet. For a better understanding of the wider issues a brief introduction explains the capacity of smartphones to protect a user’s privacy and the availability of circumvention tools against state initiated blocking. A case study researches the relationships and associations between the level of the telecommunications market development, the wealth of a country, user proficiency, the affordability of mobile technology, the level of user tolerance of state-implemented content censorship, and similar privacy threats.

Online crisis communication in a post-truth Chinese society: Evidence from interdisciplinary literature

Through a thematic content analysis of 118 journal articles from interdisciplinary fields, this study reviews the research trends of online crisis communication literature in a Chinese post-truth society, where objective facts play a diminishing role, and Chinese people are losing their trust, despite the best efforts of organizations to deal with fake news in crises. Results of this study disclose some unique crisis communication strategies (e.g., strategic distraction, rumor-rebuttal, fabrication, acting cute, and content censorship) in crises of China. Influential contextual factors such as the political system, cultural values, and media features are identified as well. Theoretical and practical implications of this study are discussed, and a greater picture of online crisis communication in contexts is posited for future research.

Bermuda Triangle – Licensing, Filtering and Privileging User-Generated Content Under the New Directive on Copyright in the Digital Single Market

The new EU Directive on Copyright in the Digital Single Market creates a peculiar triangle of obligations to license, filter and privilege user-generated content (UGC). As it abolishes the traditional safe harbour for hosting in the case of copyrighted content, it may lead to the disappearance of the open, participative internet which EU citizens currently enjoy. To avoid the loss of open, democratic avenues for online content creation, national lawmakers will have to find the right amalgam of licensing and filtering obligations on the one hand, and new use privileges that offer room for user-generated content without prior authorization on the other. Against this background, the analysis sheds light on these regulatory options and their impact on freedom of expression and information in the digital environment. It demonstrates that the two options which the Digital Single Market Directive offers to alleviate the burden of strict liability for user-generated content – the licensing approach and the filtering approach – are problematic. The licensing approach imposes the task of clearing rights for the wide variety of UGC in the highly fragmented collective licensing framework in EU Member States. The filtering approach raises the spectre of content censorship following the maxims of cost and efficiency considerations of the content and platform industry. In comparison with these scenarios, the development of a broader use privilege for UGC has the advantage of avoiding an encroachment upon fundamental rights and safeguarding breathing space for open, participative online communication. If it is combined with an obligation to pay equitable remuneration for content remixes and mash-ups that do not fulfil quotation or parody standards, it also has the advantage of generating a revenue stream that benefits not only the creative industry but also individual creators. Implementing the new Directive, national lawmakers should thus consider all options: licensing, filtering and privileging UGC. A broadened copyright limitation infrastructure that supports UGC also provides important impulses for the development of filtering technology that follows a reverse logic. Instead of seeking to find protected third party content that renders UGC impermissible, a limitation-based filtering system focuses on creative user input that may justify its online dissemination despite the inclusion of protected third party material.

Government Regulation on the Flourishing Network Audio-Visual Entrepreneurship

The network audio-visual entrepreneurship in China has achieved great progress and engendered conspicuous negative externalities in the early development stage. Few studies have investigated how media entrepreneurship coordinates with government regulation and the influence of government regulation on media entrepreneurship. This study aims at investigating government regulation on the flourishing network audio-visual entrepreneurship. This study performs semi-structured interviews with 14 respondents who are experienced in government regulation of the network audio-visual sector. It is found that license management and content censorship are principal approaches to regulating entrepreneurship. The media companies have been constrained by limited government support and social resources, and therefore endeavored to legitimate their business by collaborating with Internet conglomerates. Strict rules of content censorship discourage users from producing audio-visual content, and impose restrictions on Internet companies and other producers producing and displaying audio-visual content.

Content Censorship and Council Carelessness Why the Parliament Must Safeguard the Open, Participative Web 2.0

On 25 May 2018, the EU Council adopted its position regarding the Proposal for a Directive on Copyright in the Digital Single Market (“DSM Directive”). This negotiating mandate will have a deep impact on the further legislative process. If the Council position finally prevails, the proposed new rules will affect a wide variety of modern forms of using copyrighted content in the digital environment. In particular, the Council proposes to sharpen the rules on platform liability for user-generated content (“UGC”). As a result, the current participative web 2.0 – with unprecedented user involvement and information diversity – will hardly survive in the EU. To avoid this loss of freedom of expression and information, the time has come for the Parliament to counter the proposed content filtering systems and ensure that EU citizens retain access to the wide variety of views and cultural expressions that are currently available.

Television and Content Censorship: The Impact of Violent Content on the Developmental Stages in the Personality of the Nigerian Child

Erik Erikson’s theorization of the psychosocial developmental stages of humans, reveal the fragility of children with special attention to their biological and psychosocial characterization. The adolescent age is the period where the child is adventurous, curious and eager to explore, as he becomes fascinated with the media and begins to try out behaviours and lifestyles of his opinion leaders. The Nigerian child is more or less enthralled about foreign movies, shows and musical videos characterized by violence, thuggery, gangsterism, drug abuse, obscenity, and even promiscuity which are sometimes too much information for him to handle given his developing, immature and fragile psychology which makes it easy for these contents to influence his behaviour in every stage of his personality development as he is exposed to them. This Too Much Information (TMI) seems to be a product of media imperialism as local television contents in Nigeria, camouflaged in the concept of Americanisation and westernisation and devoid of censorship, are persistently consumed by adolescents in Nigeria. Hinged on Erikson’s theory of personality developmental stages in a child and Bandura’s social learning theory, this paper employs secondary data and a critical document analysis of previous research works to objectively opine and conclude that exposure to too much information (violent contents) on television can influence the psychosocial development of the personality of the Nigerian child.

Internet Control in Five Political Regimes in Latin America

Since 2010, internet control and security policies have been increasingly enforced (Freedom House, 2016). The literature on the subject indicates that the mechanisms of internet control vary according to the type of political regimes. This study aims to verify if control of the internet in Latin America varies according to the type of prevailing political regime. Five Latin American countries were selected as representing different political regimes: Chile, Mexico, Ecuador, Venezuela, and Cuba. Freedom of expression on the internet indicators from Varieties of Democracy are reported for each case. Freedom on the Net classification is used as an additional categorical variable. Results indicate that there is a correlation between political regime (measured by polyarchy index) and internet freedom. Indeed, the more autocratic the regime is, the more first-generation internet controls are observed, including content censorship and violations of users' rights. While in hybrid regimes, second-generation controls are more commonly observed, which involve obstacles to access, without customary content blocking and network shutdowns.

Analysis of Criminal Content Censorship (Filtering) from a Jurisprudential Framework

پالایش محتوای مجرمانه (فیلترینگ) موضوع مهمی است که با هزینه کلانی در جمهوری اسلامی ایران در حال اجراست و تعیین مصادیق آن طبق قانون جرائم رایانه‌ای به کمیته‌ای در دادستانی کل کشور واگذار شده است. بحث در زمینة کارآمدی یا ناکارآمدی این شیوۀ دفاعی در مقابله با تهاجم فرهنگی و حفظ منافع سیاسی و اقتصادی کشور از جنبه‌های مختلف مطرح است. در این مقاله سعی شده است از دیدگاه فقهی و بر مبنای اصول کلی مستخرج از قرآن، روایات و دیدگاه‌های فقهای عظام در برخی زمینه‌های سنتی مرتبط با آن، ابتدا سیاست‌های کلی دین مبین اسلام در عرصه‌های چهارگانه اخلاق، دین، سیاست و اقتصاد در دنیای واقعی استخراج گردد. سپس، سیاست‌های کلی در عرصۀ مشترک دفاع از این حوزه‌های چهارگانه به دست آید و در نهایت، این سیاست­ها در فضای مجازی تبیین گردد و در خاتمه با توجه به اینکه پالایش جزئی از نظام دفاعی کشور محسوب می‌گردد، سیاست‌های کلی مشترک عرصۀ دفاع بر دنیای مجازی تطبیق داده شود و مبتنی بر آن سیاست کلان کشور در خصوص اعمال پالایش نقد و بررسی گردد.

User awareness and tolerance of privacy abuse on mobile Internet: An exploratory study

The paper presents the results of an exploratory study about the level of privacy abuse and the awareness level of users when communicating and using mobile Internet. The study looks into the relationships and associations between the telecommunications market developmental level, the wealth of a country, users’ skills, the affordability of mobile technologies, the level of user tolerance of state-mandated content censorship, and related privacy threats. The results and findings are drawn from a collection of data gathered from ten countries which have a low reputation for respecting human rights. These countries are primarily Asian or African states. Differences within the user community tolerance levels are discussed from the perspective of the key parameters which define the level of development of the information society and also the user skill levels. For a better understanding of the issue, a brief introduction explains the capacity of smartphones to ensure user privacy, and availability of the circumvention tools for smartphones.

User tolerance of privacy abuse on mobile Internet and the country level of development

This paper presents the results of an exploratory study on the awareness of users about privacy abuse when communicating and using the mobile Internet. The study researches the relationships and associations between the level of the telecommunications market development, the wealth of a country, user proficiency, the affordability of mobile technology, the level of user tolerance of state-implemented content censorship, and similar privacy threats. The results and findings were selected from a comprehensive body of data gathered from 10 countries which have a low record of respect for human rights. Most of the selected countries are Asian or African states. Differences between the tolerance levels of the various user communities are discussed in relation to the criteria which define the level of information society development and the user skill levels. For a better understanding of the wider issues a brief introduction explains the capacity of smartphones to protect a user’s privacy and the availability of circumvention tools against state initiated blocking.

Censorship and deletion practices in Chinese social media

With Twitter and Facebook blocked in China, the stream of information from Chinese domestic social media provides a case study of social media behavior under the influence of active censorship. While much work has looked at efforts to prevent access to information in China (including IP blocking of foreign Web sites or search engine filtering), we present here the first large–scale analysis of political content censorship in social media, i.e., the active deletion of messages published by individuals.
 
 In a statistical analysis of 56 million messages (212,583 of which have been deleted out of 1.3 million checked, more than 16 percent) from the domestic Chinese microblog site Sina Weibo, and 11 million Chinese–language messages from Twitter, we uncover a set a politically sensitive terms whose presence in a message leads to anomalously higher rates of deletion. We also note that the rate of message deletion is not uniform throughout the country, with messages originating in the outlying provinces of Tibet and Qinghai exhibiting much higher deletion rates than those from eastern areas like Beijing.

The IRS Framework to Investigate Intermediary Censorship in Social Media: A Case Study of Pantip.Com

Amidst the recent political uncertainty, Thailand’s Internet censorship continues to be controversial issue because the basis for filtering seems politically motivated and actual filtering practices are inconsistent. Following the enactment of computer crime law in 2007, online service providers (OSPs) or online intermediaries in the context of Web 2.0, also known as social media, are compelled to control illegal content including content that are deemed lese majeste (defaming the royal family). This situation leads to induced self-censorship of intermediaries, often resulting in overblocking or excessive removal of content to avoid violating the law. Such filtering flaw both infringes users’ freedom of expression and impedes the business of OSPs in Thailand. The framework of innovation of retrieval system (IRS) is thus developed to investigate intermediary censorship in social media. Since the legal basis and regulatory criteria related to social media appear to be non-transparent in Thailand, IRS is expected to reveal how self-censorship of intermediaries is administered. The constructed IRS will be tested on the most popular online discussion forum, Pantip.com, to disclose its censorship practices, and to explore the opportunities and challenges of the IRS. The result of this research is expected to promote users’ freedom of expression and transparency in emerging technology of social media and to expose the criteria used by OSPs in content censorship.