11 papers from NTT Laboratories have been accepted for publication for international conference "The ACM CHI Conference on Human Factors in Computing Systems (CHI) 2026"

11 papers authored by NTT Laboratories has been accepted for publication at CHI 2026 (The ACM CHI Conference on Human Factors in Computing Systems 2026), to be held in Barcelona, Spain, from 13rd to 17th April, 2026. CHI is one of the top international conferences in the field of human-computer interaction.

Abbreviated names of the laboratories:
HI: NTT Human Informatics Laboratories
SI: NTT Social Informatics Laboratories
DT: NTT Device Technology Laboratories

The outline of the main papers is as follows.

■Exploring the Effects of Different Chatbot Voice Identities on Self-Disclosure

Yamato Mogi (Kyoto University), Wataru Akahori (SI), Naomi Yamashita (Kyoto University)

Self-disclosure has been shown to support better mental health. In recent years, chatbots have gained attention as a way to encourage self-disclosure by reducing users’ concerns about being judged by others when sharing personal information. While voice-based chatbots are now widely used, the effects of chatbot voice characteristics on users’ self-disclosure are not yet well understood. In this study, we conducted a 14-day experiment with 61 participants, comparing three chatbot voice conditions: voices synthesized from the user’s own voice, a family member’s voice, and a stranger’s voice. We analyzed differences in users’ impressions of the chatbot and in their self-disclosure. The results showed that chatbots using the user’s own voice were perceived as more attractive and encouraged deeper self-disclosure over time. These findings are expected to contribute to the design of voice-based chatbot interfaces that promote sustained self-disclosure.

■Influence or Deception? Evaluating Social Suggestions with Persuasive Statements for Security and Privacy Settings

Ayako A. Hasegawa (NICT)、Takahiro Kasama (NICT)、Mitsuaki Akiyama (SI)

Configuring security and privacy (S&P) settings can be challenging for non-expert users. As a result, people often rely on persuasive cues such as social proof (e.g., what other users choose) or recommendations from experts when making decisions. While these cues can encourage users to select more protective settings, they may also be misused as deceptive patterns that steer users toward less protective options. This study examines how source-based suggestions, such as public opinion or expert advice, influence users' decisions when combined with logical persuasive statements. It also investigates how users evaluate these approaches after learning that the suggestions may have been deceptive. We conducted an online experiment with 1,433 participants in the United States. The results show that persuasive statements amplify the influence of both social proof and authority-based cues, even when they promote less protective settings. These findings highlight the importance of designing transparent and rational persuasive interfaces for S&P settings and supporting users in critically evaluating such recommendations.

■MYOLINK Esports: Exploring EMG-Based Control Interface Through Muscle Activation and Inhibition to Enable Common Gameplay Mechanics Among Players with and Without Physical Disabilities

Masato Shindo (HI), Shiina Takano (HI), Shuto Sako (Nihon University), Akihiro Miyata (Nihon University), Ryosuke Aoki (HI)

Esports have highlighted both their potential for social inclusion and the accessibility challenges faced by individuals with physical disabilities. This study introduces a novel paradigm for inclusive esports by shifting game control from traditional kinematic inputs to kinetic inputs. An EMG-based control interface enables gameplay through force regulation (e.g., muscle activation and inhibition). The aim is to explore how this interface enables common gameplay mechanics among players with and without physical disabilities. User Study 1 involved 20 able-bodied participants performing competitive esports tasks to examine how EMG-based control accuracy is influenced by movement range, such as wrist and elbow motion. User Study 2 extended the investigation to eight participants with physical disabilities to compare control accuracy between disabled and able-bodied users. The findings suggest that the interface enables common gameplay mechanics for individuals who can separately control activation and inhibition of each muscle corresponding to each EMG sensor via calibration adjustment but disability-related involuntary muscle activity and unintended co-contraction remains a major challenge for the interface.

■Bevel or Not: Identifying the Potential of Bevels for Touch Input Accuracy on Ring Device

Yuki Kubo (HI), Arinobu Niijima (HI), Yukio Koike (HI), Buntarou Shizuki (University of Tsukuba)

Smart rings can serve as a wearable platform for off-device control of nearby devices. However, their thin bands lead to a small touch surface on the ring, limiting touch input expressivity and accuracy. To address this, we investigate the effectiveness of ring shapes that use beveled surfaces in addition to the flat outer surface as input surfaces. The distinct angles of the three surfaces limit touches to the intended surfaces, improving accuracy. A first study showed that with band widths of 6 mm or less, flat rings achieved low accuracy (63.8%) in distinguishing between left and right edge touches, whereas beveled and rounded rings achieved high accuracy (90.0%, 93.3%). In a second study of nine touch gestures on 6-mm rings, beveled rings outperformed the flat rings, achieving 92.9% (sighted) and 91.8% (eyes-free) with FPR ≤ 1%. Through these investigations, we identify ring-shape design (e.g., beveled surfaces) as a promising approach to enable expressive and precise touch input on ring devices.

■RampEMS for Stroke-Stable Brush Pressure Modulation in Calligraphy

Arinobu Niijima (HI), Yixin Peng (The University of Tokyo), Chanho Park (The University of Tokyo), Takefumi Ogawa (The University of Tokyo)

Electrical muscle stimulation (EMS) has been explored in HCI to support motor activities such as sports training and musical performance. Because contraction strength increases with stimulation intensity, EMS can modulate applied force and may support fine motor tasks such as brush-pressure control in calligraphy. However, conventional step-type EMS (StepEMS) applies the target intensity abruptly, often eliciting an involuntary onset twitch that distorts stroke trajectories. We propose RampEMS, which ramps pulse width to the target level over 1 s. This onset profile suppresses sudden movements while maintaining brush-pressure modulation. In a user study, we compared RampEMS and StepEMS in a calligraphy task and measured brush-pressure change and stroke deviation. Results indicate that RampEMS enables pressure modulation while significantly reducing stroke disturbance relative to StepEMS.

■How Does Sandbagging Manifest in Esports FPS Aiming?

Shuto Sako (Nihon Univ.), Tomoki Ikeda (Nihon Univ.), Ryosuke Aoki (HI), Akihiro Miyata (Nihon Univ.)

Sandbagging–intentionally underperforming to lower expectations– undermines fairness in esports. In esports environments, where nonverbal cues are severely constrained, sandbagging cannot be expressed or observed in the same manner as in physical sports. As a result, how sandbagging manifests as concrete in-game behav- ior remains insufficiently understood. In this study, we investigate the behavioral characteristics of sandbagging through an analysis of operation logs collected from a controlled first-person shooter aiming task. Ten participants performed the task under both high- effort and sandbagging conditions, during which detailed metrics such as shot counts and hit counts were recorded. Our analysis reveals that esports players may adopt at least two distinct sand- bagging strategies. One strategy involves consistently suppressing the number of shots while largely avoiding intentional misses. The other strategy is characterized by behavioral switching across trials, alternating between suppressing shot counts and deliberately in- creasing missed shots without reducing firing frequency. This work provides foundational empirical insights into sandbagging behavior and highlights key directions for the development of automated sandbagging detection methods in esports.

■Making Sense of Core Body Temperature: How People Interpret Circadian Rhythm Feedback in Everyday Life

Masami Takahashi (SI), Osamu Saisho (SI), Takafumi Inoue (SI), Yujiro Tanaka (DT), Daichi Matsunaga (DT)

Core body temperature (CBT), the temperature inside the body, is one of the indicators that reflects where a person’s internal body clock is within the daily cycle. With the advent of non-invasive sensors, it has become increasingly feasible to measure CBT and visualize its changes in everyday settings. However, such feedback is unfamiliar to many people, and it remains unclear how they interpret, trust, and use it.
In this study, we conducted a two-week field study (N=12) using a non-invasive CBT estimation sensor and a visualization system, followed by semi-structured interviews and thematic analysis. The results revealed that when people do not sufficiently understand the meaning of the indicators or the basis of the data, they struggle to trust and use the information, making it difficult to translate insights into everyday actions.
These findings are expected to contribute to the design of interfaces that present body clock–related data in an understandable way and support behavior change in everyday life.

■Exploring the Effects of Font Legibility Modulation for Emotionally Manipulative Social Media Content Consumption

Rintaro Chujo (The University of Tokyo),　 Mitsuaki Akiyama (SI),　 Jack Jamieson (SI), Junji Watanabe (SI), Ari Hautasaari (The University of Tokyo), Takeshi Naemura (The University of Tokyo)

Routine exposure to aggressive or highly negative social media content can undermine users' digital well-being through emotional contagion. This study investigates whether modulating typeface legibility can mitigate the influence of such content. We developed a client that detects high-arousal posts and dynamically reduces the legibility of the typeface used to display them, and conducted a seven-day field study (N=6). The results show that, compared to a standard interface, reduced legibility helps users maintain psychological distance from posts and suppress impulsive reactions.
These findings are expected to contribute to the design of interfaces that support more reflective and critical engagement with social media content while preserving user autonomy.