Several studies have analyzed traffic patterns using Vehicle Detector (VD) and Global Positioning System (GPS) data. VD records the speed of vehicles passing through detectors, GPS data captures traffic speed on the roads. However, unobserved data gaps may arise due to physical malfunctions of sensors in VD data or interruptions in satellite signal reception for GPS data. Unobserved data adds complexity to the analysis and prediction of urban traffic networks. To tackle this challenge, researchers have attempted to estimate unobserved data using spatiotemporal patterns, but approaches that rely solely on past time points are inherently less reliable. In this study, we propose a Virtual network-based Diffusion Convolutional Recurrent Neural Network (V-DCRNN) for estimating unobserved speed data in urban traffic networks using a virtual network. The virtual network is created by adding nodes and edges in virtual directions based on observed nodes at intersections, thereby enhancing the traffic network. The V-DCRNN, which utilizes the diffusion convolution process in the virtual network, uses the augmented traffic network as input to predict traffic speed. Unobserved speed data is estimated based on the values of virtual nodes predicted by V-DCRNN. We evaluate the proposed V-DCRNN model through unobserved speed data estimation experiments in the urban traffic network, where we randomly mask individual nodes and entire intersections. The main contributions of this work are as follows: 1) the design of a virtual network to model additional traffic dynamics at intersections, facilitating the estimation of unobserved speed data; 2) the development of the V-DCRNN model, which leverages a self-attention mechanism to capture spatiotemporal dependencies in urban traffic networks by incorporating the virtual network as input; and 3) an evaluation of the V-DCRNN’s ability to estimate unobserved speed data in urban traffic networks with up to 20% unobserved nodes, demonstrating robust and reliable performance.

The transfer function (TF) design is crucial for enhancing the visualization quality and understanding of volume data in volume rendering. Recent research has proposed various multidimensional TFs to utilize diverse attributes extracted from volume data for controlling individual voxel rendering. Although multidimensional TFs enhance the ability to segregate data, manipulating various attributes for the rendering is cumbersome. In contrast, low-dimensional TFs are more beneficial as they are easier to manage, but separating volume data during rendering is problematic. This paper proposes a novel approach, a two-level transfer function, for rendering volume data by reducing TF dimensions. The proposed technique involves extracting multidimensional TF attributes from volume data and applying t-Stochastic Neighbor Embedding (t-SNE) to the TF attributes for dimensionality reduction. The two-level transfer function combines the classical 2D TF and t-SNE TF in the conventional direct volume rendering pipeline. The proposed approach is evaluated by comparing segments in t-SNE TF and rendering images using various volume datasets. The results of this study demonstrate that the proposed approach can effectively allow us to manipulate multidimensional attributes easily while maintaining high visualization quality in volume rendering.

ML (Machine learning) models are used to mine inconspicuous information in big data. The model and data quality influence the performance of a machine-learning model. However, it is inefficient to modify the model, which is a black box, and low-quality data tends to cause biased learning of the model. Therefore, it is crucial to improve the data quality. Different techniques have been used to improve data quality depending on the data conditions and the data quality issues. Therefore, improving data quality is time-consuming and challenging for users with insufficient knowledge of data. Visual analytics techniques have been proposed to focus on decision support to improve data quality. However, existing studies are complicated for users to consider a comprehensive DQI (Data Quality Improvement) method for generating data suitable for ML models. Also, it remains limited in that users must directly consider all combinations of DQI processes. This paper presents a novel visual analytics system that manages data quality for use in ML models. The proposed system suggests an optimal quality improvement process with visualization techniques such as heatmap, histogram, and scatter plot to support DQI.

Machine learning (ML) models are used to mine inconspicuous information in big data. The model and data quality influence the performance of a ML model. However, modifying the ML model while measuring performance is impractical, and low-quality data causes biased model training. Therefore, improving the data quality is essential. Visual analytics systems supporting DQI (Data Quality Improvement) have been proposed in the past. However, in the studies, it is difficult for users to assess comprehensive data quality improvement methods for machine learning and to determine an appropriate data quality improvement process. In this paper, we propose a novel visual analytics system for managing data quality used in machine learning models.

Cybersickness is a symptom of dizziness that occurs while experiencing Virtual Reality (VR) technology and it is presumed to occur mainly by crosstalk between the sensory and cognitive systems. However, since the sensory and cognitive systems cannot be measured objectively, it is difficult to measure cybersickness. Therefore, methodologies for measuring cybersickness have been studied in various ways. Traditional studies have collected answers to questionnaires or analyzed EEG data using machine learning algorithms. However, the system relying on the questionnaires lacks objectivity, and it is difficult to obtain highly accurate measurements with the machine learning algorithms in previous studies. In this work, we apply and compare Deep Neural Network (DNN) and Convolutional Neural Network (CNN) deep learning algorithms for objective cy-bersickness measurement from EEG data. We also propose a data preprocessing for learning and signal quality weights allowing us to achieve high performance while learning EEG data with the deep learning algorithms. Besides, we analyze video characteristics where cybersickness occurs by examining the 360 video stream segments causing cybersickness in the experiments. Finally, we draw common patterns that cause cybersickness.

Recently, VR technology is rapidly developing and attracting public attention. However, VR Sickness is a problem that is still not solved in the VR experience. The VR sickness is presumed to be caused by crosstalk between sensory and cognitive systems [1]. However, since there is no objective way to measure sensory and cognitive systems, it is difficult to measure VR sickness. In this paper, we collect EEG data while participants experience VR videos. We propose a Deep Neural Network (DNN) deep learning algorithm by measuring VR sickness through electroencephalogram (EEG) data. Experiments have been conducted to search for an appropriate EEG data preprocessing method and DNN structure suitable for the deep learning, and the accuracy of 99.12% is obtained in our study.

In recent years, people can do most of their personal tasks, such as banking on smart devices like personal computers (PCs). Especially, Malware targeted at personal information stored on mobile are hard to detect and risks from usage patterns are even more difficult. Therefore, a means for easy recognition of the problems and the smartphone usage is necessary. In this paper, we present a personal visual analytics (PVA) system for Android security risk lifelog using app permissions to recognize the risk. Our system stores the security-related personal information on the smartphone device and utilizes it to analyze the security risk lifelog. For the risk analysis, we define security risk scores based on the app and permission statistics. Then, several linked visualizations are designed to present the risk lifelog. We have collected the security lifelog data from eight Android smartphone users and analyzed their security matters. Our PVA system enables Android smartphone users to observe, mitigate the security risk, and eventually understand how Android security risk affects their lives. Moreover, we present a user study to evaluate the PVA system with user feedback.