I'm a PhD student candidate supervised by Prof. Ravin Balakrishnan in Dynamic Graphic Project (DGP) lab at Department of Computer Science, University of Toronto. My research interest is Information Visualization and Human Computer Interaction. Before coming to UofT, I received my Bachelor's Degree (with honour) in Engineering at College of Computer Science, Zhejiang University in June, 2009 (I was also affiliated with the Chu Kochen Honors College).
Learning painting persistently in my childhood gave me the invaluable magic of discovering and creating beauties of the nature. What is more, four years of undergraduate education in Computer Science not only granted me the solid foundation of knowledge and skills but also made me desire to have the same magic on computers. And I found the perfect blend was in Information Visualization and HCI.
My research falls in specific topics about designing novel interactive visualization systems supporting multi-focus visual exploration of complicated real-world datasets, developing new methods for information seeking and navigation on web, creating interaction techniques for mobile phones, and quantitatively modelling user performance on touch screens.
I'm currently on the job market. Contact me if you are intersted in my work.
Conference Full Papers and Journal Articles
Jian Zhao, Nan Cao, Zhen Wen, Yale Song, Yu-Ru Lin, Christopher Collins.
Abstract: We present FluxFlow, an interactive visual analysis system for revealing and analyzing anomalous information spreading in social media. Everyday, millions of messages are created, commented, and shared by people on social media websites, such as Twitter and Facebook. This provides valuable data for researchers and practitioners in many application domains, such as marketing, to inform decision-making. Distilling valuable social signals from the huge crowd's messages, however, is challenging, due to the heterogeneous and dynamic crowd behaviors. The challenge is rooted in data analysts' capability of discerning the anomalous information behaviors, such as the spreading of rumors or misinformation, from the rest that are more conventional patterns, such as popular topics and newsworthy events, in a timely fashion. FluxFlow incorporates advanced machine learning algorithms to detect anomalies, and offers a set of novel visualization designs for presenting the detected threads for deeper analysis. We evaluated FluxFlow with real datasets containing the Twitter feeds captured during significant events such as Hurricane Sandy. Through quantitative measurements of the algorithmic performance and qualitative interviews with domain experts, the results show that the back-end anomaly detection model is effective in identifying anomalous retweeting threads, and its front-end interactive visualizations are intuitive and useful for analysts to discover insights in data and comprehend the underlying analytical model.
Jian Zhao, R. William Soukoreff, Xiangshi Ren, Ravin Balakrishnan.
Abstract: Scrolling interaction is a common and frequent activity allowing users to browse content that is initially off-screen. With the increasing popularity of touch-sensitive devices, gesture-based scrolling interactions (e.g., finger panning and flicking) have become an important element in our daily interaction vocabulary. However, there are currently no comprehensive user performance models for scrolling tasks on touch displays. This paper presents an empirical study of user performance in scrolling tasks on touch displays. In addition to three geometrical movement parameters --- scrolling distance, display window size, and target width, we also investigate two other factors that could affect the performance, i.e., scrolling modes --- panning and flicking, and feedback techniques --- with and without distance feedback. We derive a quantitative model based on four formal assumptions that abstract the real-world scrolling tasks, which are drawn from the analysis and observations of user scrolling actions. The results of a control experiment reveal that our model generalizes well for direct-touch scrolling tasks, accommodating different movement parameters, scrolling modes and feedback techniques. Also, the supporting blocks of the model, the four basic assumptions and three important mathematical components, are validated by the experimental data. In-depth comparisons with existing models of similar tasks indicate that our model performs the best under different measurement criteria. Our work provides a theoretical foundation for modeling sophisticated scrolling actions, as well as offers insights into designing scrolling techniques for next-generation touch input devices.
Ji Wang, Jian Zhao, Sheng Guo, Chris North, Naren Ramakrishnan.
Abstract: User reviews, like those found on Yelp and Amazon, have become an important reference for decision making in daily life, for example, in dining, shopping and entertainment. However, large amounts of available reviews make the reading process tedious. Existing word cloud visualizations attempt to provide an overview. However their randomized layouts do not reveal content relationships to users. In this paper, we present ReCloud, a word cloud visualization of user reviews that arranges semantically related words as spatially proximal. We use a natural language processing technique called grammatical dependency parsing to create a semantic graph of review contents. Then, we apply a force-directed layout to the semantic graph, which generates a clustered layout of words by minimizing an energy model. Thus, ReCloud can provide users with more insight about the semantics and context of the review content. We also conducted an experiment to compare the efficiency of our method with two alternative review reading techniques: random layout word cloud and normal text-based reviews. The results showed that the proposed technique improves user performance and experience of understanding a large number of reviews.
Jian Zhao, Christopher Collins, Fanny Chevalier, Ravin Balakrishnan.
Abstract: Many datasets, such as scientific literature collections, contain multiple heterogeneous facets which derive implicit relations, as well as explicit relational references between data items. The exploration of this data is challenging not only because of large data scales but also the complexity of resource structures and semantics. In this paper, we present PivotSlice, an interactive visualization technique which provides efficient faceted browsing as well as flexible capabilities to discover data relationships. With the metaphor of direct manipulation, PivotSlice allows the user to visually and logically construct a series of dynamic queries over the data, based on a multi-focus and multi-scale tabular view that subdivides the entire dataset into several meaningful parts with customized semantics. PivotSlice further facilitates the visual exploration and sensemaking process through features including live search and integration of online data, graphical interaction histories and smoothly animated visual state transitions. We evaluated PivotSlice through a qualitative lab study with university researchers and report the findings from our observations and interviews. We also demonstrate the effectiveness of PivotSlice using a scenario of exploring a repository of information visualization literature.
Jian Zhao, Daniel Wigdor, Ravin Balakrishnan.
Abstract: Web applications designed for map exploration in local neighborhoods have become increasingly popular and important in everyday life. During the information-seeking process, users often revisit previously viewed locations, repeat earlier searches, or need to memorize or manually mark areas of interest. To facilitate rapid returns to earlier views during map exploration, we propose a novel algorithm to automatically generate map bookmarks based on a user's interaction. TrailMap, a web application based on this algorithm, is developed, providing a fluid and effective neighborhood exploration experience. A one-week study is conducted to evaluate TrailMap in users' everyday web browsing activities. Results showed that TrailMap's implicit bookmarking mechanism is efficient for map exploration and the interactive and visual nature of the tool is intuitive to users.
Jian Zhao, Fanny Chevalier, Christopher Collins, Ravin Balakrishnan.
Abstract: A discourse parser is a natural language processing system which can represent the organization of a document based on a rhetorical structure tree---one of the key data structures enabling applications such as text summarization, question answering and dialogue generation. Computational linguistics researchers currently rely on manually exploring and comparing the discourse structures to get intuitions for improving parsing algorithms. In this paper, we present DAViewer, an interactive visualization system for assisting computational linguistics researchers to explore, compare, evaluate and annotate the results of discourse parsers. An iterative user-centered design process with domain experts was conducted in the development of DAViewer. We report the results of an informal formative study of the system to better understand how the proposed visualization and interaction techniques are used in the real research environment.
Jian Zhao, Fanny Chevalier, Emmanuel Pietriga, Ravin Balakrishnan.
Abstract: Visual representations of time-series are useful for tasks such as identifying trends, patterns and anomalies in the data. Many techniques have been devised to make these visual representations more scalable, enabling the simultaneous display of multiple variables, as well as the multi-scale display of time-series of very high resolution or that span long time periods. There has been comparatively little research on how to support the more elaborate tasks associated with the exploratory visual analysis of timeseries, e.g., visualizing derived values, identifying correlations, or discovering anomalies beyond obvious outliers. Such tasks typically require deriving new time-series from the original data, trying different functions and parameters in an iterative manner. We introduce a novel visualization technique called ChronoLenses, aimed at supporting users in such exploratory tasks. ChronoLenses perform on-the-fly transformation of the data points in their focus area, tightly integrating visual analysis with user actions, and enabling the progressive construction of advanced visual analysis pipelines.
Chronolenses has been referenced as one of the most interesting papers at InfoVis 2011.
William Soukoreff, Jian Zhao, Xiangshi Ren.
Abstract: A thought experiment is proposed that reveals a difference between Fitts' index of difficulty and Shannon's entropy, in the quantification of the information content of a series of rapid aimed movements. This implies that the contemporary Shannon formulation of the index of difficulty is similar to, but not identical to, entropy. Preliminary work is reported toward developing a model that resolves the problem. Starting from first principles (information theory), a formulation for the entropy of a Fitts' law style rapid aimed movement is derived, that is similar in form to the traditional formulation. Empirical data from Fitts' 1954 paper are analysed, demonstrating that the new model fits empirical data as well as the current standard approach. The novel formulation is promising because it accurately describes human movement data, while also being derived from first principles (using information theory), thus providing insight into the underlying cause of Fitts' law.
Jian Zhao, Fanny Chevalier, Ravin Balakrishnan.
Abstract: The need for pattern discovery in long time-series data led researchers to develop interactive visualization tools and analytical algorithms for gaining insight into the data. Most of the literature on time-series data visualization either focus on a small number of tasks or a specific domain. We propose KronoMiner, a tool that embeds new interaction and visualization techniques as well as analytical capabilities for the visual exploration of time-series data. The interface design has been iteratively refined based on feedback from expert users. Qualitative evaluation with an expert user not involved in the design process indicates that our prototype is promising for further research.
Conference Short Papers
Jian Zhao, Steven Drucker, Danyel Fisher, Donald Brinkman.
Abstract: Temporal events with multiple sets of metadata attributes, i.e., facets, are ubiquitous across different domains. The capabilities of efficiently viewing and comparing events data from various perspectives are critical for revealing relationships, making hypotheses, and discovering patterns. In this paper, we present TimeSlice, an interactive faceted visualization of temporal events, which allows users to easily compare and explore timelines with different attributes on a set of facets. By directly manipulating the filtering tree, a dynamic visual representation of queries and filters in the facet space, users can simultaneously browse the focused timelines and their contexts at different levels of detail, which supports efficient navigation of multi-dimensional events data. Also presented is an initial evaluation of TimeSlice with two datasets - famous deceased people and US daily flight delays.
Abstract: As touch-sensitive devices become increasingly popular, fundamentally understanding the human performances of multi-touch gestures is critical. However, there is currently no mathematical model for interpreting such gestures. In this paper, a novel model of multi-touch interaction is derived by combining the Mahalanobis distance metric and Fitts' law. The model describes the time required to complete an object manipulation task that includes translocation, rotation, and scaling. Empirical data is reported that validates the new model (R2>0.9). Linear relationship between the difficulty and time elapsed is revealed indicating that the model can provide guidelines for interface designers for empirically comparing gestures and devices.
Posters, Work-in-Progress, and others
Ji Wang, Jian Zhao, Sheng Guo, Chris North.
Abstract: Extracting and presenting essential information of time-varying volumetric data is critical in many fields of sciences. This paper introduces a novel approach of identifying important aspects of the dataset under the particle filter framework in computer vision. With the view of time-varying volumes as dynamic voxels moving along time, an algorithm for computing the 3D voxel transition curves is derived. Based on the curves which characterize the local data temporal behavior, this paper also introduces several post-processing techniques to visualize important features such as curve clusters by k-means and curve variations computed from curve gradients.
|Research Intern, Microsoft Research, Redmond, Dec. 2014 - Mar. 2015|
|Research Intern, Adobe Research, San Francisco, Jun. 2014 - Sep. 2014|
|Research Intern, IBM Almaden Research Center, San Jose, Jun. 2013 - Sep. 2013|
|Research Intern, Microsoft Research, Redmond, Jun. 2011 - Sep. 2011|
Robert E. Lansdale/Okino Computer Graphics Graduate Fellowship, $2000, 2015
Wolfond Fellowship, University of Toronto, $10,000, 2012
Wolfond Scholarship, University of Toronto, $5,000, 2010, 2011
University of Toronto Art & Science Fellowship, ~$30,000/year, 2009-2014
Chinese National Scholarship, ¥8,000/year, 2006, 2008
1st Class Zhejiang University Academic Scholarship, ¥5,000/year, 2006-2008
Liang Gou, Fei Wang, Jian Zhao, Michelle Zhou.
Jian Zhao, Steven Drucker, Danyel Fisher, Donald Brinkman.
CSC108 Introduction to Computer Programming (Fall 2009, Spring 2010, Fall 2011)
CSC148 Introduction to Computer Science (Fall 2010)
CSC318 Design of Interactive Computational Media (Spring 2011, Spring 2012, Fall 2013)
CSC309 Programming on the Web (Fall 2012)
CSC428/2514 Human-Computer Interaction (Spring 2014)
Bridging Data and User with Interactive Visualization Autodesk Research, Toronto, ON, 2014.11
#FluxFlow: Visual Analysis of Anomalous Information Spreading on Social Media IEEE VIS, Paris, France, 2014.11
PEARL: An Interactive Visual Analytic Tool for Understanding Personal Emotion Style Derived from Social Media IEEE VIS, Paris, France, 2014.11
Visual Comparison of Event Sequence Data Adobe Research, San Francisco, CA, 2014.9
Visual Data Exploration: A Multi-Focus Approach UOIT, Oshawa, ON, 2013.12
Visual Analytics of Online Social Media IBM Research Almaden, San Jose, CA, 2013.9
TrailMap: Facilitating Information Seeking in a Multi-Scale Digital Map via Implicit Bookmarking ACM CHI, Paris, France, 2013.5
TrailMap: Facilitating Information Seeking in a Multi-Scale Digital Map via Implicit Bookmarking ToRCHI seminar, Toronto, ON, 2013.4
Facilitating Discourse Analysis with Interactive Visualization IEEE VisWeek, Seattle, WA, 2012.10
Exploratory Analysis of Time-series with ChronoLenses IEEE VisWeek, Providence, RI, 2011.10
TimeSlice: Interactive Faceted Browsing of Timeline Data Microsoft Research, Redmond, WA, 2011.8
KronoMiner: Using Multi-Foci Navigation for the Visual Exploration of Time-Series Data ACM CHI, Vancouver, BC, 2011.5
A Model of Multi-touch Manipulation GRAND, Vancouver, BC, 2011.5
Modeling Scrolling Interactions on Touch Screens Jilin University, China, 2010.12
KronoMiner: Using Multi-Foci Navigation for the Visual Exploration of Time-Series Data KMDI seminar, University of Toronto, 2010.9
IEEE VisWeek2010, IEEE VisWeek2011, CHI2014.
CHI 2011-2015, APCHI 2012, InfoVis/VAST 2012-2014, UIST 2013-2014, PacificVis 2014, MobileHCI 2014, TVCG.