Towards Visual Exploration of Large Temporal Datasets

EasyChair Preprint 353

Abstract

Visual analytics for time series data has received considerable attention in previous literature, and different approaches have been developed to understand the characteristics of the data and to obtain meaningful information. Visualizing, analyzing and presenting large temporal datasets are important tasks to understand, navigate and explore such data. Onedimensional time-series charts are usually used to visualize time series data but if the dataset contains multiple time series with a large number of observations a high degree of overlap will occur which may obscure important information. This problem has become a vital challenge in many domains such as finance, biological systems, and meteorology. The need for analyzing and exploring large time-series data led researchers to develop various interactive visualization tools and analytical algorithms which aim to give insight into the data, and most of them either focus on a small number of tasks or a specific domain. We propose a visual analytics system and approach which aims to visualize, analyze, present and explore large temporal datasets. Our approach consists of three main stages which are preprocessing, dimensionality reduction, and visual exploration. It assists with finding the interesting features in the data which are often obscured in the line chart or the visual compression that is required to render the large datasets on a small screen. Also, it helps to obtain an overview of the entire dataset and track changes over time. Moreover, it enables the user to detect clusters and outliers and observe the transitions between data. The juxtaposed views are used to visualize and interact both with raw time series data and projection data. Different time series datasets are deployed on our system, and we demonstrate the utility and evaluate the results using a case study with two different datasets which show the effectiveness of our system.

Keyphrases: 2D Projection, Clusters, Exploration, PCA, Time series data, Time series graphs, visual analytics

@booklet{EasyChair:353,
  author    = {Mohammed Ali and Mark W. Jones and Xianghua Xie and Mark Williams},
  title     = {Towards Visual Exploration of Large Temporal Datasets},
  doi       = {10.29007/dv1j},
  howpublished = {EasyChair Preprint 353},
  year      = {EasyChair, 2018}}