Yingcai Wu | Research

My research mainly focuses on two important aspects of visual analytics of large data: trustworthy visual analytics to ensure the reliability and trustworthiness of the visual analytics, and visual analytics applications to solve practical problems. In brief, my research interests include

Visual Analytics Applications: User Behavior, Social Media, and Text Analysis

Visual Behavior Analytics
Visual Social Analytics
Visual Text Analytics

Trustworthy Visual Analytics: Methodologies and Techniques

Uncertainty-Aware Visual Analytics
Unambiguous Volume Data Visualization
Faithful Information Visualization

Visual Analytics Applications: User Behavior, Social Media, and Text Analysis

The value of visual analytics primarily lies in its capability for seamlessly combining visualization and data mining techniques to solve complex data analytics problems in practice. I have worked with domain experts from a variety of fields such as physicians and radiologists from hospitals, professors from hospitality research and hotel management, professors from media and communication studies, scientists from national labs, data scientists, software development engineers, and project managers from IT companies. The rich experience in collaboration with a wide range of domain experts enables me to gain a better understanding of the grand challenges and opportunities of visual analytics. It does not only allows me to conduct fundamental research on visual analytics techniques, but also allows me to work on multi-discipline research with domain experts. My work on visual analytics applications includes visual behavior analytics, visual social analytics, and visual text analytics.

Authors associated with * are/were the interns I have supervised

Visual Behavior Analytics

My research on visual behavior analytics focuses on explicit behavior analytics and implicit behavior analytics. Visualization of explicit user behaviors is to design effective, fundamental techniques and methodologies to visualize explicit user behaviors. StoryFlow and LoyalTracker are two systems that focus on visually and intuitively represent dynamic large-scale user behaviors.

LoyalTracker: Visualizing Loyalty Dynamics in Search Engines
Conglei Shi*, Yingcai Wu, Shixia Liu, Hong Zhou, Huamin Qu.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE VAST 2014). Honorable Mention Award, Best Paper Candidate.

StoryFlow: Tracking the Evolution of Stories
Shixia Liu, Yingcai Wu, Enxun Wei, Mengchen Liu, Yang Liu.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE InfoVis 2013)

Visual analytics of implicit user behaviors aims to design model-driven, interactive visual analytics methods to understand implicit user behaviors. OpinionFlow and OpinionSeer are two systems that combine data mining algorithms and interactive visualization techniques to reveal user behaviors hidden in the data.

OpinionFlow: Visual Analysis of Opinion Diffusion on Social Media
Yingcai Wu, Shixia Liu, Kai Yan*, Mengchen Liu, Fangzhao Wu
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE VAST 2014)

OpinionSeer: Interactive Visualization of Online Hotel Customer Feedback.
Yingcai Wu, Furu Wei, Shixia Liu, Norman Au, Weiwei Zhou, Hong Zhou, and Huamin Qu.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE InfoVis 2010)

Visual Social Analytics

Visual Social Analytics, i.e., Social Flow, studies how information flows and spreads on social meida. For example, my work of EvoRiver studies information diffusion through topic competition and topic cooperation on social media by leveraging computational models and interactive visualization.

EvoRiver: Visual Analysis of Topic Coopetition on Social Media
Guodao Sun*, Yingcai Wu, Shixia Liu, Tai-Quan Peng, Jonathan J.H. Zhu, Ronghua Liang
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE VAST 2014)

Visual Analysis of Topic Competition on Social Media
Panpan Xu*, Yingcai Wu, Enxun Wei, Taiquan Peng, Shixia Liu, Jianhua Zhu, Huamin Qu.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE VAST 2013)

Breaking News on Twitter?
Mengdie Hu, Shixia Liu, Furu Wei, Yingcai Wu, John Stasko, Kwan-Liu Ma.
In Proceedings of ACM SIGCHI 2012.
This work has been reported by MSNBC; Washington Times; ReadWriteWeb; LA Times; Poynter; GT news.

Visual Text Analytics

My work on visual text analytics focuses on interactive visualization of a large number of documents. We propose and design "semantic-preserving word clouds" that maintain semantic relationships among keywords in compact word clouds. In OpinionFlow and OpinionSeer, we combine text mining techniques and visualization approaches to detect, explore, and analyze user opinions from a large number of documents.

Semantic-Preserving Word Cloud Generation by Seam Carving.
Yingcai Wu, Thomas Provan, Furu Wei, Shixia Liu, Kwan-Liu Ma.
Computer Graphics Forum (Proceedings of EuroVis 2011).

Context-preserving Dynamic Word Cloud Visualization.
Weiwei Cui, Yingcai Wu, Shixia Liu, Furu Wei, Michelle X. Zhou, and Huamin Qu.
IEEE Computer Graphics and Applications, 30(6):42-53, 2010.

Context preserving Dynamic Word Cloud Visualization.
Weiwei Cui, Yingcai Wu, Shixia Liu, Furu Wei, Michelle X. Zhou, and Huamin Qu.
In IEEE Pacific Visualization Symposium 2010.

Trustworthy Visual Analytics: Methodologies and Techniques

Trustworthy visual analytics can be crucial for making a proper decision to avoid undesired consequences in many scenarios such as medical diagnosis and business intelligence. Data analysis, especially for exploring large data, is often an iterative, complex process where ambiguity and uncertainty can easily arise in any step and spread through the subsequent process. To resolve ambiguity and reveal uncertainty in interactive visual analytics, we need a new set of techniques and tools beyond conventional statistical analysis. Aiming at this goal, I have conducted in-depth research on the methodologies and techniques in trustworthy visual analytics including unambiguous visual analytics of volume data, faithful Information visualization, and uncertainty-aware visual analytics.

Uncertainty-Aware Visual Analytics.

Visualizing Flow of Uncertainty through Analytical Processes
Yingcai Wu, Guo-Xun Yuan, Kwan-Liu Ma.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE InfoVis 2012)

Unambiguous Volume Data Visualization.

Perceptually Based Depth-Ordering Enhancement for Direct Volume Rendering
Lin Zheng, Yingcai Wu, Kwan-Liu Ma
IEEE Transactions on Visualization and Computer Graphics, 19(3):446-459, 2013.
(Selected for presentation in IEEE VIS 2013)

Quantitative effectiveness measures for direct volume rendered images.
Yingcai Wu, Huamin Qu, Ka-Kei Chung, and Ming-Yuen Chan.
In IEEE Pacific Visualization Symposium 2010.

Perception-based transparency optimization for direct volume rendering. Ming-Yuen Chan, Yingcai Wu, Wai-Ho Mak, Wei Chen, and Huamin Qu. IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE Visualization 2009).
Honorable Mention Award, Best Paper Candidate.

Relation-aware volume exploration pipeline.
Ming-Yuen Chan, Huamin Qu, Ka-Kei Chung, Wai-Ho Mak, and Yingcai Wu.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE Visualization 2008)

Quality enhancement of direct volume rendered images.
Ming-Yuen Chan, Yingcai Wu, and Huamin Qu.
In IEEE/EG International Symposium on Volume Graphics, pages 25-32. 2007. Cover Image

Faithful Information Visualization.

ViSizer - A Visualization Resizing Framework.
Yingcai Wu, Xiaotong Liu, Shixia Liu, Kwan-Liu Ma
IEEE Transactions on Visualization and Computer Graphics, 19(2):278-290, 2013.
(Selected for presentation in IEEE VisWeek 2012)

Interactive visual optimization and analysis for RFID benchmarking.
Yingcai Wu, Ka-Kei Chung, Huamin Qu, Xiaoru Yuan, and S.C. Cheung.
IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE Visualization 2009)

Research Statement