Real-Time Exploration of Large Spatiotemporal Datasets Based on Order Statistics

Pahins, Cícero A. L.; Ferreira, Nivan; Comba, João Luiz Dihl

doi:10.1109/tvcg.2019.2914446

Cited by 9 publications

(13 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the main applications of these techniques is to represent uncertainty. Depicting uncertainty is essential for decision-making [17] and at the core of recently proposed techniques to support interactive visualization of large data [4][5][6]. The general solutions to visualize probability distributions (often sample distributions) are static representations such as error bars, confidence intervals and box plots.…”

Section: Visual Analysis Of Probability Distributions: Non-geographic...mentioning

confidence: 99%

“…Such approaches present major drawbacks. Not only can they be ambiguous, in the sense that many distributions have the same visual representation [6], but they also are often hard to interpret, even for trained individuals [25]. As a result, these techniques are designed to show the presence of variability in the distribution (i.e., uncertainty), but do not directly support the quantification of uncertainty for analytical tasks [8].…”

Section: Visual Analysis Of Probability Distributions: Non-geographic...mentioning

confidence: 99%

“…Studying probability distributions is a central task in (visual) data analysis. They are used to model uncertain behaviors present in natural phenomena [1][2][3], to build predictive models [4], and to describe the behavior/variability present in collections of data, often associated with the processes of sampling and/or aggregation [5,6]. However, effectively visualizing probability distributions presents a big challenge for visualization designers.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Comparative Study of Methods for the Visualization of Probability Distributions of Geographical Data

Srabanti

Veiga

Silva

et al. 2022

MTI

Self Cite

View full text Add to dashboard Cite

Probability distributions are omnipresent in data analysis. They are often used to model the natural uncertainty present in real phenomena or to describe the properties of a data set. Designing efficient visual metaphors to convey probability distributions is, however, a difficult problem. This fact is especially true for geographical data, where conveying the spatial context constrains the design space. While many different alternatives have been proposed to solve this problem, they focus on representing data variability. However, they are not designed to support spatial analytical tasks involving probability quantification. The present work aims to adapt recent non-spatial approaches to the geographical context, in order to support probability quantification tasks. We also present a user study that compares the efficiency of these approaches in terms of both accuracy and usability.

show abstract

Section: Visual Analysis Of Probability Distributions: Non-geographic...mentioning

confidence: 99%

Section: Visual Analysis Of Probability Distributions: Non-geographic...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Comparative Study of Methods for the Visualization of Probability Distributions of Geographical Data

Srabanti

Veiga

Silva

et al. 2022

MTI

Self Cite

View full text Add to dashboard Cite

show abstract

“…The interactive techniques can be considered an evolution of the previous ones because they make use of techniques like choropleth but with the aid of a cursor for example that allows the user to carry out some interactions. [5] use a cursor to allow the quantification of uncertainty using quantiles of the distribution or CDF of each distribution. However, the authors do not make an assessment of how This approach is effective.…”

Section: A Viewing Data Distributionsmentioning

confidence: 99%

“…These summaries are efficient in reducing the amount of data to be analyzed, however, it is possible to have completely different data collections with the same statistics as mean and variance [4]. This not only makes representation ambiguous, but it can also hide important patterns present in the data [5]. Furthermore, and most importantly, these representations fail to be well interpreted, even by individuals with statistical training [6] [7] [8].…”

Section: Introductionmentioning

confidence: 99%

A Comparative Study of Techniques of Visualization of Distributions for Geographical Data

Silva¹,

Ferreira²,

Miranda³

2020

Anais Estendidos Da Conference on Graphics, Patterns and Images (SIBRAPI Estendido 2020)

Self Cite

View full text Add to dashboard Cite

Currently, technological advances have revolutionized the way natural phenomena are studied. Natural phenomena can be represented through distributions of geographic data that are a rich source of information and can be explored in different ways. One of them is the representation of uncertainty through the distribution of probability. Modeling the uncertainty of this type of distribution and representing it in geographic visualization is complicated because maps (common types of geographic visualization) need the visual environment to represent geographic space and there are not many ways to represent any other information. One of the ways often used as a solution is statistical summarization such as mean, but summarizing the data alone may can hide the data’s behavior and generates ambiguity. The concealment of the uncertainty of the data in visualization can be justified by the way the uncertainty is represented that may not be understood by the user. Technical proposals have been proposed to represent distributions, but generally they only represent the presence and spread of uncertainty recently others approaches based on probability of proportion of data, animation and interaction have proposed to make quantification of probability, but have not been used or compared formally for geographic data. The objective was qualitatively compare main approaches to visualize probability distributions on a geographical scenario (includes factors such as distance, size and variation), using the recent proposed approaches in the context of abstract data, analytical tasks and user study. The results show which approach has the better performance in the presented cases.

show abstract

A hybrid prediction and search approach for flexible and efficient exploration of big data

Sun

Lei

et al. 2022

J Vis

View full text Add to dashboard Cite

Real-Time Exploration of Large Spatiotemporal Datasets Based on Order Statistics

Cited by 9 publications

References 32 publications

A Comparative Study of Methods for the Visualization of Probability Distributions of Geographical Data

A Comparative Study of Methods for the Visualization of Probability Distributions of Geographical Data

A Comparative Study of Techniques of Visualization of Distributions for Geographical Data

A hybrid prediction and search approach for flexible and efficient exploration of big data

Contact Info

Product

Resources

About