Bystander intervention, or upstanding, is an effective antidote to cyberbullying, but entails many challenges (e.g. self-efficacy, not knowing what to do or how to upstand). Through two studies, this paper investigates collaborative upstanding, examining how a conversational partner (human or AI) can guide bystanders through these challenges in-situ. In a paired role-play study (n=24), we found that bystanders faced significant challenges in how to intervene. Even after deciding to act, how-to challenges often reignited doubts about their self-efficacy and responsibility. Using these insights, we designed ConCUR, a chatbot that (1) encourages bystanders to co-author an upstanding message, leading them to confront how-to challenges sooner, and (2) addresses how-to challenges simultaneously with other challenges that are introduced through a flexible process. Our second study (n=20) suggests such a chatbot is effective in promoting upstanding behavior in the lab setting. We discuss the implications of in-situ collaborative upstanding to upstanding education research, framing upstanding as an iterative and flexible process rather than sequential.

We present an experimental study that investigates how LLM-driven conversational AI tools might be weaponized to facilitate, exacerbate, or commoditize coercive control. Inspired by speculative design, we construct four scenarios that combine well-known coercive control tactics with the current capabilities of conversational AI tools. Then, we explore these scenarios via interactions with popular AI agents (ChatGPT, Gemini). We find that although AI tools refuse straightforward requests for harmful content, their guardrails can be circumvented via strategies such as gradual persuasion, splitting conversations, pre-prompting, and manipulating the AI agent’s settings. Collectively, these strategies enable AI agents to be leveraged in ways that facilitate harassment, intimidation, gaslighting, monitoring, surveillance, and other coercive control tactics. To make these tools safer for everyone, we discuss opportunities for AI agents to resist being abused for coercive control via analysis of users’ conversational patterns, and ensuring that pre-programmed settings are clearly visible to prevent covert manipulation.

Ever since the introduction of internet technologies in South Korea, digital sexual violence (DSV) has been a persistent and pervasive problem. Evolving alongside digital technologies, the severity and scale of violence have grown consistently, leading to widespread public concern. In this paper, we present four eras of image-based DSV in South Korea, spanning from the early internet era of the 1990s to the deepfake scandals in the mid-2020s. Drawing from media coverage, legal documents, and academic literature, we elucidate forms and characteristics of DSV cases in each era, tracing how entrenched misogyny is reconfigured and amplified through evolving technologies, alongside shifting legislative measures. Taking a genealogical approach to read prominent cases of different eras, our analysis identifies three constitutive and interconnected dimensions of DSV: (1) the homo-social fabrication of “obscenity”, wherein victims’ imagery becomes collectively framed as obscene through participatory practices in male-dominant networks; (2) the increasing imperceptibility of violence, as technologies foreclose victims’ ability to perceive harm; and (3) the commercialization of abuse through decentralized economic infrastructures. We suggest future directions for CSCW research, and further reflect on the value of the genealogical method in enabling non-linear understanding of DSV as dynamically evolving sociotechnical configurations of harm. Warning: This paper includes descriptions and mentions of sexual violence, including violence towards minors and children.

Celebrities have used their imitated voices in media content for various purposes to entertain audiences. However, it is unclear how their fans perceive media content using synthetic celebrity voices. Fans are dedicated to celebrities and actively communicate with them on social media, sometimes resulting in the establishment of parasocial relationships. Focusing on simulated voice messages that mimic celebrities' communication behaviors, we explored how fans form parasocial relationships through synthetic media content. We conducted semi-structured interviews with 15 fans who used AI Private Call, an AI-based voice message service that employed K-Pop idols' synthetic voices. Findings show that the nature of synthetic voices could not provide impressive experiences related to their idols. However, participants believe that simulated voice messages can help make up for the limited opportunities to interact directly with idols. Based on the theoretical framework of parasocial relationship development, participants engaged with simulated voice messages, exhibiting various cognitive and behavioral actions that dynamically advanced the relationship. These behaviors include aligning voice content with idols' vocal characteristics, adhering to social norms in interactions, and confirming idols' involvement in content creation to demonstrate a sense of authenticity. Additionally, their communities associated with fandoms affect their experiences throughout the process. Based on the findings, we discuss how simulated voices might impact the relationship between fans and celebrities, highlighting the responsibility of communities and platforms. We also propose design guidelines for creating simulated celebrity voice content that fans can entertain.

Online harassment, especially networked harassment at scale, has become an increasingly serious issue that pervades many social media platforms. In this study, we investigated the nature and harms of networked harassment on Twitter through design workshops (n = 11) and developed a set of design goals focusing on empowering the individual to fight back against harassment. We designed Re:SPect, an anti-harassment tool promoting scalable and active responses to networked harassment. We evaluated Re:SPect through a simulated scenario-based study with Twitter users (n = 18) who had directly or indirectly experienced networked harassment. Our findings reveal that users felt safer and more empowered as Re:SPect enabled them to manage interactions with a larger audience. Users felt less anxious about the potential of being harassed, while the summarization features of Re:SPect allowed users to perceive the situation more objectively. Based on the findings, we discuss how Re:SPect's features could be utilized in promoting healthier online discussion, as well as theoretical implications in designing such systems.

Webtoon is a type of digital comics read online where readers can leave comments to share their thoughts on the story. While it has experienced a surge in popularity internationally, people with visual impairments cannot enjoy webtoon with the lack of an accessible format. While traditional image description practices can be adopted, resulting descriptions cannot preserve webtoons’ unique values such as control over the reading pace and social engagement through comments. To improve the webtoon reading experience for BLV users, we propose Cocomix, an interactive webtoon reader that leverages comments into the design of novel webtoon interactions. Since comments can identify story highlights and provide additional context, we designed a system that provides 1) comments-based adaptive descriptions with selective access to details and 2) panel-anchored comments for easy access to relevant descriptive comments. Our evaluation (N=12) showed that Cocomix users could adapt the description for various needs and better utilize comments.

Reflecting on stress-related data is critical in addressing one’s mental health. Personal Informatics (PI) systems augmented by algorithms and sensors have become popular ways to help users collect and reflect on data about stress. While prediction algorithms in the PI systems are mainly for diagnostic purposes, few studies examine how the explainability of algorithmic prediction can support user-driven self-insight. To this end, we developed MindScope, an algorithm-assisted stress management system that determines user stress levels and explains how the stress level was computed based on the user’s everyday activities captured by a smartphone. In a 25-day field study conducted with 36 college students, the prediction and explanation supported self-reflection, a process to re-establish preconceptions about stress by identifying stress patterns and recalling past stress levels and patterns that led to coping planning. We discuss the implications of exploiting prediction algorithms that facilitate user-driven retrospection in PI systems.

Calling out, a phenomenon where people publicly broadcast their critiques of someone to a larger audience using, has become increasingly common on social media. However, there has been concerns that it could develop into harassment, deteriorating the quality of public discourse by over-punishing individuals for minor transgressions. To investigate this phenomenon, we interviewed 32 Twitter users who had been called out, had called out, or had witnessed a calling out on Twitter. We found that a key determining factor that distinguishes criticism from harassment was the callee's ability to respond to or engage with the criticism, and that different stakeholders hold different perspectives toward how online harassment is defined. We also discovered that the distinction between callers and callees was not absolute, and that there was high interchangeability of roles both within and across events. Through these findings, we discuss design implications for the platform in promoting healthy discourse while preventing toxic behavior on social media.

Enabling the free and equal exchange of arguments on social issues in a respectful manner is an integral part of establishing the democratic ideal. However, the current manifestation of online spaces tends to facilitate the gathering of like-minded people, leading to the polarization of opinions. Such polarization inhibits the sharing of diverse opinions and deteriorates respect for disagreeing opinions. To tackle this issue, we present StarryThoughts, an online system that supports users to express and explore diverse perspectives on social issues. The system supports three types of exploration of the collected arguments online: navigating opinions based on the demographic identities of the posters, checking the the stereotypes users hold towards demographics in relation to given social issues, and engaging with opinions with semantically different point-of-views. By deploying the system to the public in co-operation with a nationwide broadcasting company, we collected 1,950 opinions with 144 free-form responses from 1,209 visitors as initial data and iterated on the design. Results from a user study with 56 participants showed that the system enables participants to explore a wide range of opinions on social issues, be more informed on the various arguments, and be more confident about their opinions. From our findings, we provide several design considerations for building online systems for supporting users to explore diverse opinions on social issues.

We examine the capacity of cooperative MIMO networks, specifically power allocation methods at the relay. Our principle finding is that power allocation using SVD and waterfilling at the source and the relay cannot achieve maximum capacity in cooperative MIMO networks. We find the optimal power allocation at the relay for the two antenna case when waterfilling is used for the direct path. A modified waterfilling power allocation method at the relay is used to increase the overall cooperative capacity when the relay is close to the source. The proposed power allocation method achieves high capacity gain in cooperative MIMO networks using a simple method when the relay has no channel information concerning the direct path. As the distance between the source and the relay increases, the outage probability at the relay plays an important role in cooperative networks. To improve the outage probability at the relay, we use a waterfilling method for the channel between the source and the relay. Finally, we compare the performance of five power allocation techniques for the multiple antenna case in terms of outage probability