Ruixuan Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ruixuan Sun has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Joseph A. Konstan
University of Minnesota, Minneapolis, MN, USA
8 shared publications - Ruoyan Kong
Department of Computer Science & Engineering, University of Minnesota, Minneapolis, Minnesota, USA
5 shared publications - Avinash Akella
IBM, New York, NY, USA
2 shared publications - Charles Chuankai Zhang
University of Minnesota - Twin Cities, Minneapolis, Minnesota, USA
2 shared publications - Chen Chen
Department of Computer Science and Engineering, University of Minnesota - Twin Cities, United States
2 shared publications - Qiao Jin
North Carolina State University, Raleigh, North Carolina, USA
2 shared publications
Latest Publications
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1719-1730
Beyond Exposure Diversity: Debiasing News Consumption With Topic-Locality Calibration and Personalized Preview Nudges
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1719-1730
Open AccessBeyond Exposure Diversity: Debiasing News Consumption With Topic-Locality Calibration and Personalized Preview Nudges
- Open Access
- 7 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 424-428
Productively Wrong: How Auditing Open Profiles Enhances Interest Awareness and Reflection in Movie Recommenders
Open User Models (OUM) leverage natural language to help users scrutinize their interests, moving beyond opaque behavioral logs to establish a ground truth of user preference. Building on this foundation, we introduce an LLM-generated open profile for a live movie recommender system that presents editable, personalized interest summaries. Unlike static profiles, this design invites users to directly inspect, audit, and reflect on the system’s inferences. In an eight-week field deployment with 1,775 active users, we found that participants actively engaged in preference construction, specifically pruning inaccurate dislikes and refining nuanced tastes. Our results show that this profile auditing process encourages high-efficiency maintenance behavior, leading to significantly increased reflection activity and engagement. We advocate for a Productively Wrong design strategy, where exposing mild AI inaccuracies serves as a catalyst to stimulate user intervention, fostering transparent systems where users actively curate their digital identity.
- 7 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 424-428
Open AccessProductively Wrong: How Auditing Open Profiles Enhances Interest Awareness and Reflection in Movie Recommenders
Open User Models (OUM) leverage natural language to help users scrutinize their interests, moving beyond opaque behavioral logs to establish a ground truth of user preference. Building on this foundation, we introduce an LLM-generated open profile for a live movie recommender system that presents editable, personalized interest summaries. Unlike static profiles, this design invites users to directly inspect, audit, and reflect on the system’s inferences. In an eight-week field deployment with 1,775 active users, we found that participants actively engaged in preference construction, specifically pruning inaccurate dislikes and refining nuanced tastes. Our results show that this profile auditing process encourages high-efficiency maintenance behavior, leading to significantly increased reflection activity and engagement. We advocate for a Productively Wrong design strategy, where exposing mild AI inaccuracies serves as a catalyst to stimulate user intervention, fostering transparent systems where users actively curate their digital identity.
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Multi-Prompting Scenario-based Movie Recommendation with Large Language Models: Real User Case Study
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessMulti-Prompting Scenario-based Movie Recommendation with Large Language Models: Real User Case Study
- Open Access
- 24 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Why They Come And Go: A Case Study of Productive Flyby Users and Their Rating Integrity Challenge in Movie Recommenders
- 24 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessWhy They Come And Go: A Case Study of Productive Flyby Users and Their Rating Integrity Challenge in Movie Recommenders
- Open Access
- 7 November 2024
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Human-Computer Interaction
- Vol. 8 (CSCW2) , 1-28
Supporting Organizations in Improving Employee Bulk E-mail --- A Tool Design and Evaluation Study
Organizations often send bulk emails to employees to make them aware of policy changes, organization plans, and events. Many of these emails, however, are long digests with many separate messages that waste employees' time and reduce their awareness. This study introduces CommTool--a prototype tool to help organizational communicators better understand their emails' performance and cost. We first interviewed 5 communicators and identified the need to measure the performance of each message within bulk email. Then we iteratively designed and deployed an organizational bulk email evaluation platform (CommTool), which enables communicators to get diverse message-level metrics such as reading time, relevance rate, comments, etc. We evaluated these designs through a 2-month field deployment with 5 communicators and 149 organization employees. We found that 1) the message-level metrics, such as reading time and relevance rate, helped communicators understand their audience and design bulk emails; 2) the cost and reputation metrics did not influence the organization leaders' decisions. We summarize with suggestions on designing organizational bulk email evaluation platforms that provide message-level performance and cost information.
- 7 November 2024
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Human-Computer Interaction
- Vol. 8 (CSCW2) , 1-28
Open AccessSupporting Organizations in Improving Employee Bulk E-mail --- A Tool Design and Evaluation Study
Organizations often send bulk emails to employees to make them aware of policy changes, organization plans, and events. Many of these emails, however, are long digests with many separate messages that waste employees' time and reduce their awareness. This study introduces CommTool--a prototype tool to help organizational communicators better understand their emails' performance and cost. We first interviewed 5 communicators and identified the need to measure the performance of each message within bulk email. Then we iteratively designed and deployed an organizational bulk email evaluation platform (CommTool), which enables communicators to get diverse message-level metrics such as reading time, relevance rate, comments, etc. We evaluated these designs through a 2-month field deployment with 5 communicators and 149 organization employees. We found that 1) the message-level metrics, such as reading time and relevance rate, helped communicators understand their audience and design bulk emails; 2) the cost and reputation metrics did not influence the organization leaders' decisions. We summarize with suggestions on designing organizational bulk email evaluation platforms that provide message-level performance and cost information.
- Open Access
- 8 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1308-1313
AI-based Human-Centered Recommender Systems: Empirical Experiments and Research Infrastructure
This is a dissertation plan built around human-centered empirical experiments evaluating recommender systems (RecSys). We see this as an important research theme since many AI-based RecSys algorithmic studies lack real human assessment. Therefore, we do not know how they work in the wild that only human experiments can tell us. We split this extended abstract into two parts – 1) A series of individual studies focusing on open questions about different human values or recommendation algorithms. Our completed works include user control over content diversity, user appreciation on DL-RecSys algorithms, and human-LLMRec interaction study. We also propose three future works to understand news recommendation depolarization, personalized news podcast, and interactive user representation; 2) An experimentation infrastructure named POPROX. As a personalized news recommendation platform, it aims to support the longitudinal study needs from the general AI and RecSys research community.
- 8 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1308-1313
Open AccessAI-based Human-Centered Recommender Systems: Empirical Experiments and Research Infrastructure
This is a dissertation plan built around human-centered empirical experiments evaluating recommender systems (RecSys). We see this as an important research theme since many AI-based RecSys algorithmic studies lack real human assessment. Therefore, we do not know how they work in the wild that only human experiments can tell us. We split this extended abstract into two parts – 1) A series of individual studies focusing on open questions about different human values or recommendation algorithms. Our completed works include user control over content diversity, user appreciation on DL-RecSys algorithms, and human-LLMRec interaction study. We also propose three future works to understand news recommendation depolarization, personalized news podcast, and interactive user representation; 2) An experimentation infrastructure named POPROX. As a personalized news recommendation platform, it aims to support the longitudinal study needs from the general AI and RecSys research community.
- 5 October 2023
- Published by Taylor & Francis in International Journal of Human–Computer Interaction
- Vol. 40 (22) , 7233-7247
Interactive Content Diversity and User Exploration in Online Movie Recommenders: A Field Experiment
- 5 October 2023
- Published by Taylor & Francis in International Journal of Human–Computer Interaction
- Vol. 40 (22) , 7233-7247
Interactive Content Diversity and User Exploration in Online Movie Recommenders: A Field Experiment
- Open Access
- 16 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 291-295
Less Can Be More: Exploring Population Rating Dispositions with Partitioned Models in Recommender Systems
In this study, we partition users by rating disposition - looking first at their percentage of negative ratings, and then at the general use of the rating scale. We hypothesize that users with different rating dispositions may use the recommender system differently and therefore the agreement with their past ratings may be less predictive of the future agreement. We use data from a large movie rating website to explore whether users should be grouped by disposition, focusing on identifying their various rating distributions that may hurt recommender effectiveness. We find that such partitioning not only improves computational efficiency but also improves top-k performance and predictive accuracy. Though such effects are largest for the user-based KNN CF, smaller for item-based KNN CF, and smallest for latent factor algorithms such as SVD.
- 16 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 291-295
Open AccessLess Can Be More: Exploring Population Rating Dispositions with Partitioned Models in Recommender Systems
In this study, we partition users by rating disposition - looking first at their percentage of negative ratings, and then at the general use of the rating scale. We hypothesize that users with different rating dispositions may use the recommender system differently and therefore the agreement with their past ratings may be less predictive of the future agreement. We use data from a large movie rating website to explore whether users should be grouped by disposition, focusing on identifying their various rating distributions that may hurt recommender effectiveness. We find that such partitioning not only improves computational efficiency but also improves top-k performance and predictive accuracy. Though such effects are largest for the user-based KNN CF, smaller for item-based KNN CF, and smallest for latent factor algorithms such as SVD.
- Open Access
- 30 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Getting the Most from Eye-Tracking: User-Interaction Based Reading Region Estimation Dataset and Models
A single digital newsletter usually contains many messages (regions). Users’ reading time spent on, and read level (skip/skim/read-in-detail) of each message is important for platforms to understand their users’ interests, personalize their contents, and make recommendations. Based on accurate but expensive-to-collect eyetracker-recorded data, we built models that predict per-region reading time based on easy-to-collect Javascript browser tracking data. With eye-tracking, we collected 200k ground-truth datapoints on participants reading news on browsers. Then we trained machine learning and deep learning models to predict message-level reading time based on user interactions like mouse position, scrolling, and clicking. We reached 27% percentage error in reading time estimation with a two-tower neural network based on user interactions only, against the eye-tracking ground truth data, while the heuristic baselines have around 46% percentage error. We also discovered the benefits of replacing per-session models with per-timestamp models, and adding user pattern features. We concluded with suggestions on developing message-level reading estimation techniques based on available data.
- 30 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessGetting the Most from Eye-Tracking: User-Interaction Based Reading Region Estimation Dataset and Models
A single digital newsletter usually contains many messages (regions). Users’ reading time spent on, and read level (skip/skim/read-in-detail) of each message is important for platforms to understand their users’ interests, personalize their contents, and make recommendations. Based on accurate but expensive-to-collect eyetracker-recorded data, we built models that predict per-region reading time based on easy-to-collect Javascript browser tracking data. With eye-tracking, we collected 200k ground-truth datapoints on participants reading news on browsers. Then we trained machine learning and deep learning models to predict message-level reading time based on user interactions like mouse position, scrolling, and clicking. We reached 27% percentage error in reading time estimation with a two-tower neural network based on user interactions only, against the eye-tracking ground truth data, while the heuristic baselines have around 46% percentage error. We also discovered the benefits of replacing per-session models with per-timestamp models, and adding user pattern features. We concluded with suggestions on developing message-level reading estimation techniques based on available data.
- Open Access
- 19 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-18
Collaborative Online Learning with VR Video: Roles of Collaborative Tools and Shared Video Control
Virtual Reality (VR) has a noteworthy educational potential by providing immersive and collaborative environments. As an alternative but cost-effective way of delivering realistic environments in VR, using 360-degree videos in immersive VR (VR videos) received more attention. Although many studies reported positive learning experiences with VR videos, little is known about how collaborative learning performs on VR video viewing systems. In this study, we implemented two collaborative VR video viewing modes based on the way of group video control, synchronized or shared (Sync mode) and non-synchronized or individual (Non-sync mode) video control, against a conventional VR video viewing setting (Basic mode). We conducted a within-subject study (N = 54) in a lab-simulated remote learning environment. Our results show that collaborative VR video modes (Sync and Non-sync mode) improve users’ learning experiences and collaboration quality, especially with shared video control. Our findings provide directions for designing and employing collaborative VR video tools in online learning environments.
- 19 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-18
Open AccessCollaborative Online Learning with VR Video: Roles of Collaborative Tools and Shared Video Control
Virtual Reality (VR) has a noteworthy educational potential by providing immersive and collaborative environments. As an alternative but cost-effective way of delivering realistic environments in VR, using 360-degree videos in immersive VR (VR videos) received more attention. Although many studies reported positive learning experiences with VR videos, little is known about how collaborative learning performs on VR video viewing systems. In this study, we implemented two collaborative VR video viewing modes based on the way of group video control, synchronized or shared (Sync mode) and non-synchronized or individual (Non-sync mode) video control, against a conventional VR video viewing setting (Basic mode). We conducted a within-subject study (N = 54) in a lab-simulated remote learning environment. Our results show that collaborative VR video modes (Sync and Non-sync mode) improve users’ learning experiences and collaboration quality, especially with shared video control. Our findings provide directions for designing and employing collaborative VR video tools in online learning environments.