591 research outputs found
Replication Data for: "Russian Invasion of Ukraine and Chinese Public Support for War"
No full textReplication Data for: "Russian Invasion of Ukraine and Chinese Public Support for War
Jie gou xing tu you hua de wu cha jie xian: li lun ji ying yong
No full textZhou, Zirui.Thesis Ph.D. Chinese University of Hong Kong 2015.Includes bibliographical references (leaves 94-102).Abstracts also in Chinese.Title from PDF title page (viewed on 02, November, 2016).Zhou, Zirui
Mitigating Regional Accent Bias in ASR Systems
No full textEnd-to-end Automatic Speech Recognition (ASR) systems improved drastically in recent years and they work extremely well on many large datasets. However, research shows that these models failed to capture the variability in speech production and have biases against the variant caused by the regional accented speech. Moreover, ASR research on regional accents is primarily done in languages used by a large population, like English and Arabic, and the effect of regional accented speech on E2E ASR systems in non-popular languages is still unknown. It is important to know the effect of regional accented speech on E2E ASR systems as it helps researchers to build an inclusive E2E ASR system. In this project, I aim to mitigate the biases against regional accented speech. I select standard speech and regional accented speech from CommonVoice's French and German datasets. I combine the state-of-the-art Conformer Recurrent Neural Network Transducer model with Multi-Domain Adversarial Training (MDAT) to boost the performance of regional accented speech while not hurting the performance of the standard speech. Moreover, since the regional accented speech is typically low-resourced, I study the amount of data required for effective MDAT, as well as the effect of different domain classifiers on the performance of Multi-Domain Adversarial Training. Experimental results show that MDAT can mitigate the biases against regional accented speech in both French and German. The best model in French reduces the bias by around 12% and the best model in German reduces the bias by around 7%. Additionally, MDAT is an effective method for bias mitigation as it can achieve similar performance as the MDAT model trained with the full dataset using only a small amount (e.g. 30 minutes) of untranscribed regional accented speech. Finally, different domain classifier architectures were found to have similar effects on the results of MDAT, thus there is no significant differences among the domain classifier in this project.Electrical Engineering | Embedded System
Essays on Censorship and Public Opinion in Authoritarian Regimes
No full textConventional wisdom has long regarded censorship as a top-down, repressive tool for authoritarian governments to suppress political criticism and maintain regime stability, and therefore unpopular among the public. This dissertation proposes a novel normalization theory and presents various pieces of empirical evidence to challenge these conventional understandings of censorship and public opinion in authoritarian regimes. In the first chapter, I introduce the normalization theory and claim that the Chinese public does not necessarily perceive online censorship as repressive but as a normal part of Internet governance. Drawing from around 28 million censored posts on social media and two survey experiments, I demonstrate that, in addition to politically threatening content, non-political content is also censored on a substantial scale, which subsequently increases public support for censorship. In the second chapter, I further investigate why the public perceives censorship as normal by analyzing public participation in the censorship process. I propose a novel bottom-up perspective on censorship and demonstrate that censorship participation leads to higher public support for the censorship apparatus. In the third chapter, I challenge the conventional wisdom that authoritarian censorship tends to target positive exposure of foreign liberal democracies. Using a novel dataset of Chinese social media articles about liberal democracies from 2018 to 2022, I show that impeding mass exposure to democratic institutions rather than defaming the West is the primary strategy of Chinese censorship. This study underscores the Chinese regime\u27s lingering insecurity about public knowledge of liberal democratic systems. Taken together, this dissertation highlights how the normalization of censorship, ordinary citizens\u27 participation, and strategic censorship of information about democracy contribute to maintaining public support and sustaining regime survival in autocracies
Participatory Censorship: Experimental Study
No full textContrary to the conventional top-down view of government censorship, ordinary citizens in authoritarian regimes frequently participate in censorship by submitting censorship demands. Yet, existing research overlooks such a bottom-up perspective of censorship. This study provides the first empirical description of public demand for political censorship and analyzes its consequences on public opinion toward government censorship in authoritarian regimes. I argue that public participation in censorship shifts citizens' perception of censorship and increases their support for the regime's censorship policy. To causally test the relationship between participation and censorship support, I design an original experiment using custom-engineered, simulated social media pages. The experiment will be conducted in the summer of 2022
Public Support for the Use of Force in Authoritarian Regimes
No full textThe Russian invasion of Ukraine has significantly increased concerns over authoritarian military aggression around the world. The concerns over the potential spread of authoritarian military aggression raise important questions about the determinants of authoritarian military aggression and the effectiveness of potential measures, such as economic sanctions, to limit such aggression. Specifically, we ask: Does the Russian invasion of Ukraine increase support for the Chinese government's use of military force in international affairs among the Chinese public? Do Western economic sanctions and military (in)actions against the Russian invasion of Ukraine affect such public support for the use of military force in China?
We plan to conduct an online survey experiment in mainland China. The purpose of the survey experiment is to present respondents with relevant information about the invasion and test the effect of learning about the Russian invasion of Ukraine on Chinese individuals' attitudes toward the Chinese government's use of force
Normalization of Censorship: Evidence from China
No full textCensorship in China is massive in scale and still growing. Previous studies of censorship in China focus primarily on government criticism and collective actions and find that political censorship will lead to backlash from the citizens. However, novel observational evidence shows that a majority of censored articles are unrelated to either government criticism or collective actions. This study investigates the implication of such expansion in range and scale of censorship. It hypothesizes that when censorship is expanded beyond government criticism and collective action, backlash from the citizens will be reduced
Normalization of Censorship: Follow-up Study Pre-Analysis Plan
No full textThis is a follow-up study of the project ``Normalization of Censorship.'' In this project, I ask why a majority of the Chinese public support government censorship. I argue that the range of censorship targets extends beyond politically threatening content such as government criticism and collective action. A substantial proportion of censorship in China targets seemingly harmless non-political content such as entertainment, culture, and advertisement. As a consequence, public perception of censorship shifts. When citizens frequently encounter censorship of non-political content, they are less likely to view censorship as suppression of political opposition. Rather, censorship becomes viewed as a normal government policy that regulates both political speeches and non-political content.
The follow-up study has two objectives: (1) to replicate the results in the previous experiment and provide support for the main hypotheses; (2) to provide support for the additional hypotheses and distinguish the normalization theory from alternative explanations. To test these hypotheses, I plan to conduct an online survey experiment. In the experiment, I will ask respondents to read some social media posts during which I manipulate the censorship signal they receive. After that, I will measure their support for the censorship apparatus, the regime, censorship of political content, and censorship of non-political content
Demanding Censorship
No full textTo the surprise of many international observers, ordinary Chinese citizens frequently demand censorship from the government for both political and non-political reasons. Why do many Chinese citizens support the government's censorship policy? How does public participation in the censorship process contribute to citizens' support for censorship? This project aims to analyze the effect of public participation in censorship on public support for government censorship
Doublespeak: The Effectiveness of China's Hard and Soft Propaganda during Policy Changes
No full textHow effective are different authoritarian propaganda strategies in cultivating support and maintaining social stability during periods of crisis and policy changes? Existing literature has identified two main strategies of authoritarian propaganda: soft propaganda which uses persuasion to convince the population that the government is competent, and hard propaganda which uses dry and threatening language to signal regime strength and deter potential rebellions. Yet, few have directly compared the effectiveness of these different strategies, especially during periods of crisis and policy changes when support and social stability matter most. In this study, we theorize that both hard and soft propaganda strategies have pitfalls, particularly during policy changes, that are previously overlooked. For hard propaganda, bland slogans do not just signal regime strength but also government policy resolve, which might increase, rather than decrease, collective action because citizens are more likely to participate in rightful resistance against local authorities. For soft propaganda, policy changes force the regime to send contradictory arguments that undermine their persuasiveness. To support our arguments, we leverage the rare opportunity of China's COVID policy U-turns in November and December 2022. We propose to conduct an original survey experiment in China that randomly exposes respondents to different COVID propaganda messages ranging from hard to soft and pro-restriction to pro-reopening. We then measure their assessment of China's COVID responses as well as their willingness to participate in protests against COVID policies
- …
