(prefix indicating ordinal number, e.g. first, number two etc)11 arrive at (place or time)IEEE Pacific Visualization Symposium (PacificVis 2018) Held at Kobe University, Japan, on April 10. Many experts and scholars in the field of visualization were invited to report, and several students from the 360 Enterprise Security Visualization Lab were also invited to participate in the exchange, so please ask Xiao Yueyue to bring you a forward live report today.
sessions (prefix indicating ordinal number, e.g. first, number two etc) One day forPacific VAST workshops， There are four elementsTechnical Session and a thematic report（Key Note）。
Session 1: Biomedical Applications
The first report is entitled "A Visual Analytics System to Support the Formation of aHypothesis from Calcium Waves Data". The mechanism of calcium wave formation is not yet clear, and biology experts, through collaboration with visualization experts from Kyoto University, have found that it would be helpful to solve this problem if we could obtain an overview of calcium wave propagation and understand the mechanism of calcium wave propagation in narrow regions, but existing visualization methods often fail to meet both conditions. Therefore, an Overview-Detail-Interaction style visual analysis system is proposed to understand and explore the calcium wave data obtained by the biologists.
(prefix indicating ordinal number, e.g. first, number two etc) The second report is entitled“MultiSciView: Multivariate ScientificX-ray Image Visual Exploration with Cross-Data Space Views”。X Rays are a complex data， It encodes multiple properties of the detectors， It's an effective way of perceiving our world， But scientists currently don't have effective tools to tap into its hidden value。 therefore， Visualization experts from the United States have come up with a technique calledMultiSciView The visual exploration system of the。 This is a multi-view interactive visualization system， Enables collaborative exploration of images and their associated attributes and feature spaces。
The third presentation is entitled "An Uncertainty-aware Workflow for Keyhole Surgery Planning usingHierarchical Image Semantics". The importance of minimally invasive surgery in daily clinical practice cannot be overstated, and visualization experts from Germany have proposed a visualization workflow based on the concept of hierarchical image semantics to help clinicians examine and compare different surgical paths and the extent to which patients are affected. The workflow, which includes a linked view system to indicate the location of selected surgical paths and how these paths affect the patient, has been applied to minimally invasive surgical planning for brain tumor resection and knee surgery.
Key Note: ExplainableMachine Learning
After a wonderful presentation in the morning, the afternoon started with a keynote presentation by Ms. Liu Shixia from Tsinghua University. Machine learning is very effective in solving real-world problems such as AI and data mining, but due to its complex functionality and working mechanism, it is still a "BlackBox" in most people's eyes, and if its models are not well understood, high-performance models will require repeated time-consuming trial-and-error processes. This presentation shows the main challenges of interactive machine learning and exemplifies the solutions with several visual analytics techniques and examples.
Session2: Text Analysis
The first presentation is entitled "ageLens: A Visual Analytics System to Support Multifaceted Exploration ofMOOC Forum Discussions". How useful are MOOC forums? Those who have used them know this, but the high volume and heterogeneity of MOOC forum content makes it challenging to effectively analyze forum data from different discussion perspectives and to integrate the various messages into a coherent understanding of issues of concern. Professor Xiaolong Zhang from Penn State University shared a visual analytics system called MessageLens, which is used to analyze the rich content of communication within forums and the social dynamics of forum users.
The second presentation is entitled "Metro-Wordle: An Interactive Visualization for Urban TextDistributions based on Wordle". With the growth of cities and the explosion of information, POI analysis becomes challenging. Visualization experts from Shanghai Jiao Tong University provide a visual design that combines metro maps and text to extract useful information from POI data and uncover the spatial distribution of text. If you happen to be in Shanghai and you happen to be a foodie, their Case Study may provide you with a food map of Shanghai!
Session3: VA Design Tools and Systems
(prefix indicating ordinal number, e.g. first, number two etc) A report entitled“TideGrapher: Visual Analytics of Tactical Situations for RugbyMatches”。 For all you rugby lovers out there， Visualization experts from Keio University in Japan have come up with a technique called“TideGrapher” The visual analysis system of the， Used for visual analysis of tactics in rugby。 In this visual analysis system， Pay special attention to the lateral position of the ball， Specialized spatial substrates designed according to the spatio-temporal trajectory of the ball， and provides a set of basic interactions， at the same time， Innovative parameters are proposed for quantitative analysis。
The second presentation is entitled "VisComposer: A Visual Programmable Composition Environment forInformation Visualization". Visualization experts from Zhejiang University have proposed a visualization development environment called "VisComposer" that supports expressive visualization development using a drag-and-drop visualization interface, effectively balancing the gap between software and visualization experts and making visualization building more flexible.
Session4: VA System and Design Study
The first report is entitled "LongLine: Visual Analytics System for Large-scale Audit Logs". Audit logs record raw events in the operating system and are of high interest to security professionals and system administrators. However, the complexity and size of audit logs increases in real time, and the difficulty of understanding and analyzing them is significant. Visualization experts from Seoul National University in Korea have proposed an interactive visual analysis system for large-scale audit logs called "LongLine".
The second presentation is entitled "A Visual Analytics System for Optimizing the Performance of Large-ScaleNetworks in Supercomputing Systems". Visualization experts from UC Davis have proposed a visual analysis system, based on a network of Dragonfly topologies, to investigate temporal behavior and optimize the communication performance of supercomputers. Combining interaction visualization with timing analysis methods while using multi-view collaboration helps to analyze behavior patterns on different parallel applications, system configurations, and different hierarchical network structures.
Our field partners will continue to share live content from the IEEE Pacific Visualization Symposium (PacificVis 2018) conference, so stay tuned!
Well, that's all for today~
Feel free to discuss and interact with us in the comments too!
Please stay tuned to our public
We'll keep sharing more interesting and dry stuff~