China
Africa
Explore chapters and articles related to this topic
Introduction to IT Transformation of Safety and Risk Management Systems
Published in Qamar Mahboob, Enrico Zio, Handbook of RAMS in Railway Systems, 2018
Coen van Gulijk, Miguel Figueres-Esteban, Peter Hughes, Andrei Loukianov
The term visual analytics arose around 2005, being defined as a combination of “… automated analysis techniques with interactive visualizations for an effective understanding, reasoning and decision making on the basis of very large and complex data sets” (Keim et al. 2008). VA is a multidisciplinary area that attempts to obtain insight from massive, incomplete, inconsistent, and conflicting data in order to support data analysis, but usually requires human judgment (Thomas and Cook 2005). Visual analytics explicitly combines four dimensions of big data (volume, variety, veracity, and value). Implicitly, the last attribute, velocity, may be considered a computational requirement to take into account in performing analysis and decision making. Therefore, it might be said that visual analytics is a variant of ‘big data analytics’ supported by interactive visualization techniques (Figueres-Esteban, Hughes, and Van Gulijk 2015). Visual Analytics has yielded five pillars where visualization can support data analysis tasks, viz., data management, data analysis/mining, risk communication, human–computer interaction, and information/scientific visualization (Figure 35.5). In the scope of this paper, data analysis is most relevant.
Data visualization
Published in Benjamin S. Baumer, Daniel T. Kaplan, Nicholas J. Horton, Modern Data Science with R, 2021
Benjamin S. Baumer, Daniel T. Kaplan, Nicholas J. Horton
Thus far our discussion of data visualization has been limited to static, two-dimensional data graphics. However, there are many additional ways to visualize data. While Chapter 3 focuses on static data graphics, Chapter 14 presents several cutting-edge tools for making interactive data visualizations. Even more broadly, the field of visual analytics is concerned with the science behind building interactive visual interfaces that enhance one's ability to reason about data.
Introduction
Published in Alexandru Telea, Data Visualization, 2014
Recognizing the need to combine visualization solutions with data analysis and data mining front-ends, a new discipline has emerged from the information visualization, scientific visualization, and data-mining communities: visual analytics. Briefly put, the central goal of visual analytics is to provide techniques and tools that support end users in their analytical reasoning by means of interactive visual interfaces [Wong and Thomas 04, Thomas and Cook 05, Meyer et al. 12]. Although, at the current moment, no clearly defined boundary exists between visual analytics and the more traditional infovis and scivis fields it has emerged from, several aspects differentiate visual analytics from its predecessors. First, visual analytics focuses on the entire so-called sensemaking process that starts with data acquisition, continues through a number of repeated and refined visualization scenarios (where interaction is heavily involved to allow users to explore different viewpoints or test and refine different hypotheses), and ends by presenting the insight acquired by the users on the underlying phenomena of interest. As such, visual analytics is typically characterized by a tight combination of data analysis, data mining, and visualization technologies and tools. Separately, visual analytics typically focuses on processes or datasets which are either too large, or too complex, to be fully understood by a single (static) image. As such, data mining, multiple views, and interactive and iterative visual inspection of the data are inseparable components of visual analytics. Figure 1.2(d) shows a snapshot from a visual analytics process. Here, the repository 2D view used can be seen as a classical infovis example. However, to understand, or make sense, of the development process taking place during the history of the studied software repository, several sorting and coloring options, interactively chosen by the user, need to be applied.
Harnessing the Visual Salience Effect With Augmented Reality to Enhance Relevant Information and to Impair Distracting Information
Published in International Journal of Human–Computer Interaction, 2023
Xin Lei, Yueh-Lin Tsai, Pei-Luen Patrick Rau
Situated analytics draws from the two domains of visual analytics and AR to support a new form of in-situ interactive visual analysis. First, visual analytics is a multidisciplinary research area spanning analytical reasoning with visualization, and its basic idea is to visually represent information to enhance the interpretation of the information for better decision-making (Keim et al., 2008). Second, AR is a useful tool for visualization and can enrich the physical worldview by overlaying virtual information directly on top of physical objects in the real world (Thomas et al., 2014; White & Feiner, 2009a). According to Thomas et al. (2018), situated analytics employs data representations that are organized concerning germane objects, places, and persons for understanding, sensemaking, and decision-making.
‘Breaking’ news: uncovering sense-breaking patterns in social media crisis communication during the 2017 Manchester bombing
Published in Behaviour & Information Technology, 2020
Our subset of data encompasses a total of 708,147 Twitter postings, including original tweets, retweets (569,124), and replies. We converted the dataset to a graph representation, which is composed of 292,451 nodes (unique users/participants) and 569,124 edges (retweets). Accordingly, 139,023 tweets during the 6-hour period are of unique content, including replies. Figure 2 displays the retweet frequency per minute, which mirrors the content sharing activities of the Twitter community, synchronous to the crisis events near the Manchester arena. To visualise this bulk of data, we employed the visual analytics tool Tableau.2 A peak of information exchange can be observed at around 00:43 UTC with 108 retweets per minute. The activity decreases in the following hours, presumably due to local night time and the clarification of pressing issues, i.e. basic information about the crisis have been published and validated.
An interactive platform for the analysis of landscape patterns: a cloud-based parallel approach
Published in Annals of GIS, 2019
Jing Deng, Michael R. Desjardins, Eric M. Delmelle
Effectively communicating the results of landscape metric studies at fine spatial and space-time scales across a large extent can also be challenging. A variety of studies address the issue of visualizing large amounts of spatial and space-time data (Gong et al. 2013; Delmelle et al. 2014; Desjardins et al. 2018a; Desjardins et al. 2018b; Desjardins et al. 2018a); but the figures are static, which may occlude key patterns. Interactive visual analytics can facilitate the discovery of patterns and mitigate the challenge of communicating large volumes of information (Szewrański et al. 2017). Interactive visualization platforms may allow users to zoom, pan; select and compare layers and attributes in one or more windows; improve analysis with supplemental graphs, plots, and tables; and other supplemental functionality (Andrienko and Andrienko 1999; MacEachren et al. 2004; Chertov et al. 2005; Kulawiak et al. 2010; Schumann and Tominski 2011; Kinkeldey 2014; Andrienko et al. 2016; Schiewe 2018).