BIGDATA: F: Critical Visualization Technologies for Analyzing and Understanding Big Network Data

BIGDATA：F：分析和理解大网络数据的关键可视化技术

• 批准号: 1741536
• 负责人: Kwan-Liu Ma
• 金额: $ 56万
• 依托单位: University of California-Davis
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-10-01 至 2022-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1741536&HistoricalAwards=false
• 关键词: BIGDATA; Critical; Visualization; Technologies; Analyzing

Big data presents both opportunities and challenges to all fields of study and practice. Visualization has been proven effective as a knowledge discovery and storytelling tool for big data. This project aims to develop new visualization technologies for big network data that will both illustrate empirical findings and generate new discoveries. Although many network visualization techniques and tools have been introduced, visualizing large, dynamic networks to extract key entities, structures, and trends from the network data remains a challenging task. Most of the existing network visualization solutions were not designed for handling dynamic networks and are too slow for interactive exploration of large networks. This project will closely examine the integral parts of a holistic solution for the problem of big network visualization. The research study will be largely driven by the data analysis needs of sociological studies such as finding hidden associations between multiple networks; however, the project team will also investigate the solution's applicability in areas such as emergency management, life science and cyber security. The resulting technologies are expected to drastically enhance one's ability to explore and understand large, complex dynamic networks for knowledge discovery, critical decision making, and storytelling. This research effort is timely because of the explosive growth of data and common use of graphs as both the internal data structure and a visual representation in data-driven applications. Those who must deal with large, complex dynamic network data for their work will benefit from the advanced visualization technologies resulted from this research project. Students participating in this project will acquire strong interdisciplinary research skills for real-world problem solving. This research underscores the importance of providing a comprehensive solution to the understanding of big data containing complex relations, structure, and trends. Primary research topics are: (1) Visual depiction and exploration of big network data; (2) Modeling and visualizing dynamic network data; (3) Visual monitoring and analysis of live, streaming network data; and (4) Provenance and storytelling with dynamic network data. This project will explore and integrate new network modeling, reduction, and visualization techniques for analyzing large, multivariate dynamic graphs. The resulting research innovations will both enhance existing methods and investigate new approaches to dynamic network visual analytics and drastically improve their usability for real-world applications. The targeted applications, emergency service and sociology, present the project team with some of the most challenging problems to address in making sense of heterogeneous dynamic big networks data. The collaborating domain experts are fully committed to participating in the evaluation work, which promises to produce usable technologies that will enable respondents to look at the data in new ways and uncover intricate relations among different entities/events for critical decision making and mitigation planning. The project results will be disseminated to the visualization community and beyond through annual conferences, workshops, and tutorials, and also through the project website which will include project status updates and resulting images, videos, and prototype software.

大数据为所有研究和实践领域带来了机会和挑战。可视化已被证明是有效的大数据知识发现和讲故事工具。该项目旨在为大型网络数据开发新的可视化技术，既可以说明经验发现并产生新的发现。尽管已经引入了许多网络可视化技术和工具，但可视化大型动态网络以从网络数据中提取关键实体，结构和趋势仍然是一项艰巨的任务。大多数现有的网络可视化解决方案不是为处理动态网络而设计的，并且对于大型网络的交互式探索太慢。该项目将仔细研究整体解决方案的整体部分，以解决大型网络可视化问题。 研究研究将主要由社会学研究的数据分析需求驱动，例如在多个网络之间找到隐藏的关联。但是，项目团队还将调查解决方案在紧急管理，生命科学和网络安全等领域的适用性。预计所得的技术将大大提高人们探索和理解知识发现，关键决策和讲故事的大型，复杂的动态网络的能力。这项研究工作之所以及时，是因为数据的爆炸性增长和图形的共同使用既是内部数据结构，又是数据驱动的应用程序中的视觉表示。那些必须处理大型，复杂的动态网络数据的人将受益于该研究项目所产生的先进可视化技术。参加该项目的学生将获得强大的跨学科研究技能，以解决现实世界中的问题。这项研究强调了为理解包含复杂关系，结构和趋势的大数据提供全面解决方案的重要性。主要研究主题是：（1）大型网络数据的视觉描述和探索； （2）建模和可视化动态网络数据； （3）现场流媒体网络数据的视觉监视和分析； （4）使用动态网络数据的出处和讲故事。 该项目将探索和集成新的网络建模，还原和可视化技术，以分析大型多元动态图。由此产生的研究创新将增强现有方法，并研究动态网络视觉分析的新方法，并大大提高其对现实应用程序的可用性。有针对性的应用程序，紧急服务和社会学，向项目团队介绍了一些最具挑战性的问题，以解决异质动态大型网络数据。合作的领域专家完全致力于参与评估工作，该工作有望生产可用的技术，使受访者能够以新的方式查看数据，并发现不同实体/事件之间的复杂关系，以进行关键的决策和缓解计划。该项目结果将通过年度会议，研讨会和教程将可视化社区以及超越，还将通过项目网站传播，其中包括项目状态更新以及结果图像，视频和原型软件。

项目成果

Staged Animation Strategies for Online Dynamic Networks

• DOI: 10.1109/tvcg.2020.3030385
• 发表时间: 2020-09
• 期刊: IEEE Transactions on Visualization and Computer Graphics
• 影响因子: 5.2
• 作者: Tarik Crnovrsanin;Shilpika;Senthil K. Chandrasegaran;K. Ma

A visual analytics framework for reviewing multivariate time-series data with dimensionality reduction

• DOI: 10.1109/tvcg.2020.3028889
• 发表时间: 2020-01-01
• 期刊: arXiv preprint arXiv:2008.01645
• 作者: Fujiwara, T.;Sakamoto, N.;Ma, K. L
• 通讯作者: Ma, K. L

BC tree-based spectral sampling for big complex network visualization

• DOI: 10.1007/s41109-021-00405-3
• 发表时间: 2021-08-21
• 期刊: APPLIED NETWORK SCIENCE
• 影响因子: 2.2
• 作者: Hu, Jingming;Chu, Tuan Tran;Ma, Kwan-Liu
• 通讯作者: Ma, Kwan-Liu

Supporting Analysis of Dimensionality Reduction Results with Contrastive Learning

• DOI: 10.1109/tvcg.2019.2934251
• 发表时间: 2020-01-01
• 期刊: IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
• 影响因子: 5.2
• 作者: Fujiwara, Takanori;Kwon, Oh-Hyun;Ma, Kwan-Liu
• 通讯作者: Ma, Kwan-Liu

Resolving Conflicting Insights in Asynchronous Collaborative Visual Analysis

解决异步协作可视化分析中的冲突见解

• DOI: 10.1111/cgf.13997
• 发表时间: 2020
• 期刊: Computer Graphics Forum
• 影响因子: 2.5
• 作者: Li, Jianping Kelvin;Xu, Shenyu;Chris) Ye, Yecong;Ma, Kwan‐Liu
• 通讯作者: Ma, Kwan‐Liu