Sage Journals: Discover world-class research

Abstract

Visualization is integral to uncovering hidden information in data and providing users with intuitive feedback for decision-making. Data visualization is crucial for transforming complex data into actionable insights across various domains. In recent years, coronavirus disease vaccines have become increasingly available to much of the population. However, the CDC (Centers for Disease Control and Prevention) often fails to consider multidimensional coronavirus pandemic data from a side-by-side perspective, limiting the ability of medical professionals and individuals to compare and interact with comprehensive data visualizations. Effectively displaying coronavirus and vaccination data collected from multiple sources is essential for interpreting pandemic transmission patterns and vaccine efficiency. This paper presents a new platform for innovative data visualizations that offers users intuitive feedback and a complete data story. We designed algorithms to seamlessly combine multiple parameters, synchronize attributes, and dynamically visualize data over time on a single webpage. Instead of integrating all attributes into a single plot, which can be overwhelming due to space limitations and make it difficult to extract crucial information from overcrowded display components, we developed algorithms to classify, enhance, and group all parameters based on their relationships and similarities. Furthermore, a side-by-side visualization method was created to dynamically link all parameters in multiple images for data exploration, trend comparison, hidden information detection, and correspondence analysis. Our platform provides real-time performance, enabling healthcare professionals to make informed decisions, communicate findings effectively, and uncover patterns that might not be apparent in raw data. The proposed multidimensional data visualization algorithms have broad applications in general data exploration and revealing hidden information.

Keywords

Visualization multidimensional data synchronization feature grouping multi-parameter dynamic linking revealing hidden information

Introduction

A highly contagious coronavirus challenged the world on scales many could not have foreseen in the spring of 2019.¹ The virus has gone from a disease in a remote region of the Wuhan Province of China to a global pandemic that has affected everyone in the world either directly or indirectly.² In the United States alone, approximately 29 million cases and nearly 500,000 deaths have arisen as a result.³ The spread of the coronavirus worldwide has sparked interest in data analytics and visualizations. Everyone wants to see and understand how case counts are rising, the virus’s impact on themselves and their community, and their role in slowing the spread.^4,5

The coronavirus pandemic outbreak provides an opportunity to gain critical information for research and developing treatment strategies. It offers a significant case study on the importance of modern public communication for spatially related data and the effective implementation of tools whose insights can impact decision-making at different levels. As Anthony Fauci et al. states, “Every outbreak provides an opportunity to gain important information.”⁶

Open data on the spread of pandemic diseases is published to make it easier for related parties such as the government, health organizations, research institutes, and the public to obtain an overview of the pandemic disease problem. However, published open data is still presented numerically, making it difficult to get visual feedback and reveal hidden information.⁷ The coronavirus pandemic has forced the world to interact with data to make decisions at the individual level.⁸ Visualization can generate an understanding of complex data and extract helpful information, playing a vital role in scientific analysis.⁹

Visualization can enhance understanding, improve engagement, and facilitate efficient communication by presenting information in a more user-friendly way. It plays a pivotal role in transforming raw data into actionable knowledge. Collecting diverse data sets from various sources enables researchers to develop robust visualizations that allow a more precise story of the data to be told.¹⁰ This empowers professionals to make informed decisions and communicate findings effectively.¹¹ Data visualization is a suitable tool to represent inherent trends in data, enhance decision-making, empower teams with human-centric situational awareness insights, and reveal hidden information, such as how each state or territory handles its vaccination rollout and response to an overworked public health system.¹² Viewing data in a format that provides intuitive feedback will enable more effective local, state, and federal decision-making.¹³

However, visualization also has disadvantages. Creating high-quality visualizations can be time-consuming, leading to potential misinterpretation. Interpretation of visualizations can be subjective, as different individuals may derive different meanings from the same visual representation, and creating overly complex visualizations can be overwhelming. Therefore, developing efficient data visualization algorithms is a challenging task.¹⁴ Failure to communicate effectively might significantly increase the possibility of delaying pandemic outbreak control, undermine public trust and compliance, and lead to unnecessarily prolonged economic, social, and political turmoil.¹⁵

In response to the recent coronavirus outbreak, several organizations and private citizens have designed and published online explanatory data visualization tools to communicate disease data supported by a spatial dimension. Agencies such as the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) have collected data from various demographics of the world’s population. This data provides a comprehensive picture of how differently the coronavirus pandemic and other diseases affect different people, enabling doctors, first responders, and everyday citizens to understand how the disease is spreading and how to combat the rising case numbers and misinformation. Moreover, several data visualizations have been created to inform the public about the rapidly evolving crisis. For example, Zhang et al.¹⁶ examined design practices in producing coronavirus data visualizations to provide a component of crisis information infrastructure. Padilla et al.¹⁷ presented methods for improving trust while maintaining trend interpretation performance through multiple forecast visualization guidelines. Latif et al.¹⁸ reviewed the rapidly growing body of recent research on various coronavirus research activities to leverage data science. Sharma et al.¹⁹ proposed a data visualization tool that includes conceptual and data-driven information.

To demonstrate the temporal and spatial evolution characteristics of the epidemic development, Jia et al.²⁰ designed and developed an epidemic simulation and visualization system based on WebGIS. John and Shaiba²¹ developed interactive maps, tables, and heatmaps to analyze coronavirus data, with maps playing a vital role in understanding the spatial orientation of infection spread. To better understand the epidemiological development in each country and obtain country-specific surveillance, a web application was built using R programing,²² which has been used to interpret and visualize coronavirus data both nationally and internationally.

For the purpose of exploring the effects of the pandemic on the population through a geographical spatiotemporal view, Maciejewski et al.²³ developed a visual analytics toolkit (PanViz) for analyzing the impact of decision measures implemented during a simulated pandemic influenza scenario. They utilized Tableau for interactive visualization and analysis of coronavirus data.²⁴ Tao et al.²⁵ reviewed existing data analysis and visualization techniques along with their applications for coronavirus disease and future pandemic management. Engledowl and Weiland²⁶ discussed plausible reasons for the misrepresentation of pandemic-related data in the media and by governmental agencies.

Some publications utilized socio-economic factors to formulate and test causal hypotheses, such as Banerjee et al.,²⁷ who advocated an explanatory framework from cooperative game theory to analyze socioeconomic disparities that are causal to the spread of COVID-19 in the United States. To find the difference in intrinsic vulnerability to coronavirus disease between countries, Kong et al.²⁸ studied the correlations of the basic reproduction number with social, economic, and environmental covariates by compiling data on predictors from seven categories, including demographics, disease, economics, environmental, habitat, health, and social for some countries. Raphael and Schneider²⁹ proposed narrative research on how the pandemic unfolded and reshaped the landscape of socioeconomic inequality in America.

Furthermore, the development of coronavirus vaccines has progressed at an unprecedented pace. Within nearly 2 years after the first lockdown was announced in the United States, multiple vaccines have been developed by various companies.^30,31 Understanding the patterns and number of cases in relation to how vaccines are distributed at the state level can provide critical insights into how state and local governments can effectively address the evolving pandemic situation.^32,33 Visualizations play a crucial role in revealing hidden patterns in the collected vaccination data and bridging gaps where data is incomplete or missing, particularly for marginalized groups or regions.^34,35

Motivation and contributions

Incorporating vaccination data into visualizations is crucial for comprehensively understanding the pandemic’s progress. Visualizing the impact of vaccinations alongside other COVID-19 metrics can help policymakers, healthcare professionals, and the public make informed decisions. However, understanding and analyzing pandemic and vaccination data can be challenging due to complex structures, multiple parameters, and evolving dynamic features, making extracting critical information for data analysis and pattern Still, these tools cannot provide the functions and flexibility of simultaneously showing the multi-parameter coronavirus data acquired from many sources over time. The visualization research has prediction challenging. The published research and existing data visualization tools such as Tableau,³⁶ Matlab,³⁷ and Domo³⁸ can display multi-parameter data, including coronavirus data. not considered the effect of vaccination, so it cannot deliver a whole picture of the pandemic progress. There are also publications about socio-economic factors on COVID-19 causal hypotheses.^27–29 However, this research usually fails to comprehensively integrate multiple parameters of coronavirus data to provide visual feedback and expectations. All these difficulties and shortcomings will delay the speed of providing scientific support for front-line medical workers and government agencies to make the most appropriate and timely decisions.^39,40

Effective data visualization plays a crucial role in helping people understand complex information. To address the challenges of multi-parameter pandemic data visualization, we first present algorithms and a software platform for data visualizations that provide users with intuitive feedback and enable them to see a complete story of where the coronavirus data is concerned. Next, we seek to explore the coronavirus data further by considering multi-dimensional variables and various parameters over time. The critical point of this paper is to dynamically visualize and analyze the coronavirus and vaccination data collected from various sources, and multiple parameters are displayed interactively to analyze the relationship between vaccination, disease, and time. Visualizing such data dynamically can provide valuable insights and help users better understand the trends and patterns in the information.

The assumptions toward developing our solution are through cleaning, processing, and analyzing the data, which can lead to valuable insights and extract hidden information from coronavirus and vaccination data. The significant contribution of this research is the development of an extended web-based platform and new algorithms to represent all data sets from various aspects in one cohesive visualization. Due to the effectiveness of color and mapping diagrams possessing the ability to convey information hidden in raw data visually,⁴¹ the visualization featured in this work is a dynamic choropleth map to show how data changes from state to state using color mapping, allowing us to visualize pandemic information tied to geography and analyze data across locations. At the same time, a multi-source synchronization scheme is developed to link the quantitative charts with a colored map to enhance all the selected graph components simultaneously. A bar chart overlay method is designed to show multi-dimensional variables in one graphic visualization. The connected scatterplot is combined with a bar chart to visually compare data acquired from multiple sources, enabling users to read the numbers for each state while also maintaining the ability to see how the disease trends overall and keep sight of both the forest and the trees as it were.

The novelty of our algorithm lies in its ability to process and explore high-dimensional data within an interactive framework effectively. Multiple attributes of pandemic data related to coronavirus disease and its impact on people in the United States are synchronized and dynamically visualized over time on a single webpage. We classified, enhanced, and grouped the data parameters based on their relationships, functions, and similarities. This information is illustrated using a side-by-side visualization method, seamlessly linking all parameters across multiple images on the web. This approach enables users to visualize how the pandemic varies across geographic areas and observe the level of variability within a region. For example, we group features by age, vaccination rate, case and death rate, and vaccination by age. While fewer than four features exist in each group, we synchronize multiple groups and dynamically link them together on a single webpage. This allows our method to explore data with very high dimensions.

We integrate textual information into our data visualization platform, dynamically displaying multiple parameters in one cohesive visualization. By aggregating diverse sets of coronavirus data and presenting them dynamically over time, our platform can aid medical researchers in gaining deeper insights into disease progression and transmission trends. When augmented with relevant socio-economic factors, enabling epidemiologists to formulate and test causal hypotheses, our platform has the potential to inspire pandemic operations experts and policymakers to devise effective strategies for detecting, controlling, and mitigating the impact of the coronavirus. As demonstrated in Figure 1, the progress of the COVID-19 pandemic transmission and its impact over time can be effectively visualized. This includes mapping the total vaccination rate, the relationship between vaccines distributed and people vaccinated, the number of people recovered and confirmed per 10,000, as well as the percentage of deaths due to coronavirus and boosters used per 100. Figure 2 illustrates the dynamic visualization results of vaccination progress, showing the number of vaccinated and boosters administered at four time points. Figure 3 demonstrates the dynamic visualization of coronavirus disease progress at four-time points.

Figure 1.

The dynamic rendering result of the progress on the covid pandemic transmission and affection over time, including full vaccination rate mapping, relationship between vaccination distributed and people vaccinated, people recovered and confirmed per 10,000, as well as the percentage of death due to coronavirus and boosters used per 100. (a) COVID-19 full vaccination rates per 100 by state; (b) vaccination distribution, vaccination, and boosted by state over time, and (c) confirmed, recovered, and death rate by state over time.

Figure 2.

Demonstration of dynamic visualization results at four time points: (a) 08/27/2021, (b) 10/29/2021, (c) 03/24/2022, and (d) 10/04/2022 –to illustrate the progression of coronavirus disease. The x-axis represents recoveries per 10,000, the y-axis shows confirmed cases per 10,000, and the circle radius corresponds to deaths per 10,000. The 52 states in the United States are grouped into five regions, each distinguished by a unique color.

Figure 3.

Illustration of four time points: (a) 11/10/2020, (b) 11/28/2021, (c) 01/09/2022, and (d) 02/23/2022 –from the dynamic visualization of the progression of coronavirus disease. The x-axis represents recoveries per 10,000, the y-axis shows confirmed cases per 10,000, and the circle radius represents deaths per 10,000. The 52 states in the United States are grouped into five regions, with each group distinguished by a unique color.

The developed platform aims to enhance risk perception by incorporating interaction as a critical component. This feature will help professionals effectively communicate the importance of pandemic data while assisting regular citizens in making informed and safe decisions in a world affected by coronavirus disease. In our next research phase, we will collect data on socio-economic factors and integrate it with the coronavirus data used in this study. We will design machine learning models to link socio-economic information with COVID-19 incidence rates and predict future development trends. Additionally, we will invite domain experts to evaluate our visualization algorithms and software platform through targeted case studies.

The organization of this paper is as follows. Section Materials and Methods introduces the platform and tools used in this research, along with the procedures for preparing and processing data. Next, it details the algorithms for data visualization and synchronization, followed by an explanation of how data rendering progresses over time. Section Results and Analysis presents the experimental results using the proposed system for multi-parameter pandemic data visualization and synchronization over time. Finally, Section Conclusion provides concluding remarks.

Materials and methods

Platform and tools used

In this paper, we focus on creating a modern, responsive, and user-friendly web application to address the new needs of the coronavirus disease outbreak. Multi-factor data about coronavirus and vaccination data over time from January 12, 2021, to May 10, 2023, are used in this research. To build a robust and visually appealing web application that works seamlessly across different platforms, cutting-edge web technologies, such as the programing language JavaScript, and the markup language HTML5 (Hypertext Markup Language 5), are used in our interactive system design for structuring and presenting content on the World Wide Web, and Cascading Style Sheets (CSS) is employed for the formatting of a document written in a markup language. Using JavaScript for interactivity, HTML5 for content structure, and CSS for styling reflects a standard and effective approach in contemporary web development.

Visual Studio Code is used for software development, in which the $d 3 . csv ()$ sends HTTP requests to the specified URL to load a .csv file or data and execute a callback function with parsed CSV data objects, and $d 3 . json ()$ sends HTTP requests to the specified URL to load JASON file and execute the callback function with parsed JSON data objects. We used JavaScript library, Data-Driven Documents (D3, which is a JavaScript library for producing dynamic, interactive data visualizations in web browsers. It makes use of Scalable Vector Graphics, HTML5, and Cascading Style Sheets standards), to produce dynamic, interactive data visualizations in web browsers. D3 makes use of Scalable Vector Graphics, HTML5, and Cascading Style Sheets (CSS) standards. Moreover, d3-scale, and d3-legend are employed to map multi-dimension data to a visual representation. Google Chrome is used as the web browser to load the visualization, and time points are used as an important parameter to show the disease development and the effect of vaccines on controlling the pandemic transmission over time.

An open-source, cross-platform JavaScript runtime environment Node.js⁴² is used to build the server that runs JavaScript on server-side scripting, which enables an asynchronous event-driven JavaScript runtime, and Node.js web application framework Express⁴³ is employed to provide features for web application and launch client-side data visualization in web browsers.

Data preparation and processing

In this project, multiple data sets are taken from the New York Times “The Coronavirus Pandemic,”⁴⁴ the State Health Facts,⁴⁵ the Centers for Disease Control and Prevention (CDC),⁴⁶ and Kaggle,⁴⁷ respectively. The collected data are stored in five corresponding CSV (comma-separated values) files, which are text files that have a specific format allowing data to be saved in a table-structured format.

The major objective of this project is to visualize pandemic and vaccination data and extract information about their connections in each state of the United States. Therefore, we selected the data for vaccination (at least one dose, fully vaccinated, booster), as well as the total COVID-19 case and death rates (Table 1). To find relationship variations in different age groups, we collected the percentage of each age group in each state of the United States (Table 2), the percentage of fully vaccinated individuals in each age group (Table 3), and the percentage of updated boosters in each age group (Table 4). For these four tables, we only selected six states as examples to show the data format and values in each category. We did not list the entire dataset used in the visualization and analysis in these tables.

Table 1.

Data sample of the first 10 states (alphabetical order) acquired from Ref.⁴⁴: vaccination at least one dose, fully vaccinated, booster, and total covid case and death rate.

State	Id	One dose	Fully vac.	Booster	Total case	Total death
Alabama	AL	64	52	20	31.2	1.4
Alaska	AK	72	64	31	41.1	0.5
Arizona	AZ	76	64	29	33.3	1.4
Arkansas	AR	69	56	24	32.6	1.3
California	CA	85	74	42	29.8	0.9
Colorado	CO	82	72	40	30.0	0.8
Connecticut	CT	95	82	49	26.2	1.2
Delaware	DE	86	72	35	32.5	1.0
Dis. of Col.	DC	95	82	38	25.2	0.8

Table 2.

Data sample of the six states from 11 to 16 (alphabetical order) acquired from Ref.⁴⁵: the percentage of age group 0–18, 19–25, 26–34, 55–64, as well as 65 and over.

State	Id	0–18	19–25	26–34	35–54	55–64	65 & over
Florida	FL	20.8	7.7	11.2	25.1	13.9	21.3
Georgia	GA	25.0	8.6	12.2	26.6	12.6	14.9
Hawaii	HA	22.6	7.1	11.5	25.2	13.2	20.4
Idaho	ID	26.1	8.6	11.7	24.8	12.1	16.7
Illinois	IL	23.5	8.5	12.1	26.2	13.2	16.6
Indiana	IN	24.6	8.9	11.9	25.1	13.1	16.4

Table 3.

Data sample of the six states from 17 to 22 (alphabetical order) acquired from Ref.⁴⁶: the percentage of fully vaccinated in the age group 5–11, 12–17, 18–64, and 65 and up.

State	Id	5–11	12–17	18–64	65 & up
Iowa	IA	27	51	68	95
Kansas	KS	27	54	70	95
Kentucky	KY	21	45	64	88
Louisiana	LA	14	40	60	87
Maine	ME	46	71	86	95
Maryland	MD	47	80	85	95

Table 4.

Data sample of the six states from 17 to 22 (alphabetical order) acquired from Ref.⁴⁶: the percentage of updated booster in the age group 5–11, 12–17, 18–64, and 65 and up.

State	Id	Under 18	18–64	65 & up
Vermont	VT	23	52	88
Virginia	VA	12	40	74
Washington	WA	13	44	77
West Virginia	WV	4	26	60
Wisconsin	WI	10	39	81
Wyoming	WY	4	22	61

We aimed to visualize and analyze the relationship between pandemic rates and vaccination over time in each state of the United States. Therefore, we collected data over time for total confirmed cases, deaths, and recoveries from 04/12/2020 to 03/07/2021 and the percentage of people vaccinated, vaccinations distributed, and boosters used from 08/27/2021 to 05/10/2023 (Table 5). This table only shows five states and three time periods as an example of the data format and value range used in this research. The collected vaccination data span over 2 years and are used to illustrate vaccination progress, providing a detailed summary with multiple metrics such as the number of vaccines given in each state, the number of people who received the first dose, the second dose, the booster percentage, and daily vaccinations per million, etc. We also display and analyze the dynamic connections between vaccination, booster shots, coronavirus disease confirmation rate, the percentage of recovery, and deaths over time.

Table 5.

Illustration of data samples of the six states in six regions of the United States acquired from Ref.⁴⁷

Total	Total	Total	Date	State	Region	Date	People	Distributed	Boosters
Confirmed	Deaths	Recovered					Vaccinated (%)	(%)	(%)
1280	27	367	04/12/2020	Arkansas	Southeast	08/27/2021	52.73	121.92	0.90
1621	70	842	04/12/2020	Minnesota	Midwest	08/27/2021	61.52	128.81	0.41
9762	315	2014	04/12/2020	Texas	Southwest	08/27/2021	56.63	130.21	3.34
2303	18	218	04/12/2020	Utah	West	08/27/2021	55.96	115.74	1.29
727	27	15	04/12/2020	Vermont	Northeast	08/27/2021	75.87	157.16	2.45
96,524	1645	86,447	10/16/2020	Arkansas	Southeast	04/30/2022	66.67	205.51	23.14
117,106	2252	104,547	10/16/2020	Minnesota	Midwest	04/30/2022	75.06	225.30	44.59
839,050	17,623	716,015	10/16/2020	Texas	Southwest	04/30/2022	73.11	213.38	31.35
90,461	529	66,683	10/16/2020	Utah	West	04/30/2022	75.23	193.59	28.52
1903	58	66,683	10/16/2020	Vermont	Northeast	04/30/2022	93.45	288.88	58.00
324,653	5297	314,732	03/07/2021	Arkansas	Southeast	05/10/2023	70.09	277.46	38.24
489,116	6614	474,175	03/07/2021	Minnesota	Midwest	05/10/2023	79.12	328.38	81.84
2,733,793	45,241	2,470,308	03/07/2021	Texas	Southwest	05/10/2023	76.99	278.71	45.28
374,438	1975	357,312	03/07/2021	Utah	West	05/10/2023	75.53	262.70	55.00
15,964	208	13,145	03/07/2021	Vermont	Northeast	05/10/2023	99.99	414.90	104.04

The data total confirmed, deaths, and recovered are in the time range from 04/12/2020 to 03/07/2021, while the data of the percentage of people vaccinated, vaccination distributed, and boosters used are from 08/27/2021 to 05/10/2023. Three time points used in this data demo.

Table 1 shows sample data of the first 10 states (alphabetical order) acquired from,⁴⁴ including the state names and corresponding IDs, as well as the population percentage of vaccination, that is, at least one dose, fully vaccinated, updated booster, and the total coronavirus disease rate and death rate at the time October 20, 2022. The data of the population distribution by age is scrapped from.⁴⁵ The data set contains the population percentage of five age groups in each state in 2021, and the six states from 11 to 16 are illustrated in Table 2. The data of coronavirus vaccinations in the United States by defined age groups is demonstrated in Tables 3 and 4, which contain six selected states to show information on fully vaccinated and updated boosters in the age group, respectively.

The example vaccination and disease cases data are shown in Table 5. The data of total confirmed cases, deaths due to coronavirus disease, and recovered are listed from 04/12/2020 to 03/07/2021 in five states, each from a different region. The percentage of vaccinated, vaccine distributed, and the percentage of boosters used between 08/27/2021 and 05/10/2023 is also listed in this table. The states in the United States are grouped with regions in unique colors, and the vaccination and disease information is visualized dynamically over time, which can be interactively rendered over time through changing time sliders on the web-based platform. The items used in our visualization include total boosters per 100, total vaccinations per 100, people vaccinated per 100, percentage of fully vaccinated, daily vaccinations, and vaccination distributed per 100. All the vaccination data is illustrated over time in more than 2 years.

The analysis framework identifies relationships among grouped features, including total confirmed cases, deaths, and recoveries. Total confirmed cases provide insights into the extent of pandemic transmission, while total deaths indicate disease severity. Recoveries help evaluate treatment effectiveness and highlight vaccine efficacy. Additionally, parameters related to vaccination rates gage the level of protection vaccines offer against health outcomes. This information is essential for public health authorities and healthcare professionals to develop effective, transparent communication strategies on vaccine safety and effectiveness. It also guides the selection of appropriate vaccines to control the pandemic effectively.

The $d 3 . csv (url, callback)$ function in D3 is used to send HTTP requests for fetching the file of type CSV at the specified URL and execute the callback function with parsed CSV data objects, which can be used as multidimensional arrays for data processing and calculation. In this project, we save the CSV files in the Node.js server’s local folder. The $d 3 . json ()$ is employed to send an HTTP request to the specified URL to load a JSON (JavaScript Object Notation) file of the United States that contains 52 states. The state name in the JSON file is applied to match the loaded CSV file with the state name as matching IDs for data processing and color mapping in each state. The $d 3 . timeParse ()$ is used to get a time series over around 2 years, which is used to show the data progress and development trends such as disease, vaccination, and recovery. The data preparation and processing are described in Figure 4.

Figure 4.

Diagram illustrating the research workflow, depicting the pipeline from synchronizing multiple raw data sources in CSV format and a JSON map of the United States through data processing and calculation to component transformation, dynamic visualization, and multi-parameter progress evaluation over time within a web browser.

We designed two types of synchronization mechanisms, as illustrated in Figure 4. The first type uses the state name as an ID to dynamically link all grouped attributes. Parameters are classified into several groups based on similarity and function, with the state name assigned as the parameter ID for each group. When users interact with a parameter widget on the graphical user interface, a signal is emitted to all the side-by-side data visualization components. Consequently, all components with the same ID are enhanced and updated based on the received signals. The second type of synchronization involves separating the data into individual time points. When users adjust the current time point using a slider, a signal is sent to all components in the web-based user interface. This triggers updating and synchronizing all parameter values to reflect the selected time point.

Data visualization and synchronization

In the developed platform, each graph chart is created on an independent SVG (Scalable Vector Graphics, which is an XML-based vector image format for defining two-dimensional graphics, having support for interactivity and animation) element, which is rendered with a corresponding JavaScript function. As shown in Figure 5, six graph charts are rendered simultaneously to display the acquired data sets listed in Tables 1 to 4. The data set columns of the percentage of vaccination, coronavirus cases, and death due to coronavirus disease are dynamically mapped to the United States map illustrated in Figure 5(a). The mapping is controlled by the menu, which is created with an HTML select element. The “As least one dose” is the default mapping, and users can change the mapping selection interactively. The five mapping graphs are displayed in Figure 6, which include “As least one dose”, “Fully Vaccinated,”“Booster Enhancement,”“Covid Case Rate,” and “Covid Death Rate.” When users change the mapping from the system menu, the SVG element is cleared, and the selected graph chart is rendered again with the corresponding new data.

Figure 5.

The platform diagram for data visualization in a web browser includes: (a) a map of the United States with states represented using different color scales, (b) state-specific age group population distribution, (c) vaccination and booster percentages, (d) COVID-19 case and death rates, and (e) vaccination distribution by age groups showing the percentage of fully vaccinated residents and (f) the percentage of residents with a booster. The state of Oklahoma is selected, highlighted, and synchronized across all visualizations.

Figure 6.

Demonstration of multi-source synchronized visualization using Texas as an example, dynamically enhanced with mulberry color, including (a) state vaccination map, (b) age group distribution, (c) vaccination rates (at least one dose, fully vaccinated, and booster), (d) COVID case and death rates, (e) vaccination percentage by age group for fully vaccinated individuals, and (f) vaccination percentage by age group for booster recipients.

The JavaScript portion of the code is contained in an HTML script element, and each data set is set to a unique ID and selected by D3. An SVG (Scalable Vector Graphics) element is appended to the ID for 2D graphics rendering. As demonstrated in Figure 5(b) and (c), our algorithm displays multiple data columns on one SVG element, which greatly improves the efficiency of space usage and quantitative information comparison. We also combine two display methods of the connected scatterplot and bar chart on one SVG element, which is described in Figure 5(d) and (e) and is used for information synchronization.

The connected scatterplot and bar chart use the same linear scale, mapping from zero to the maximum value in the current column, as selected by the second HTML select element. The mapped values are used in the y-axis, and the state names are extracted by their IDs and displayed on the x-axis using $d 3 . scaleBand$ to lay out the state’s IDs along the x-axis range. The data map uses a linear scale of mapping the minimum and maximum values of the current column to a color range changing from yellow to blue. A linear scale is a type of measurement or representation in which the distances between consecutive points or values are equal. The scale progresses in a straight, constant manner. On this scale, a blue color $ represents the largest value, while a yellow color $rgb (255, 255, 120)$ represents the smallest value. All the chart graphs are rendered in their individual JavaScript functions.

To create the connected scatterplot and bar chart graphs, the data are used to calculate the width of each rectangle, and the index of each value in the data set is used to calculate the y attribute. The largest values are shown at the top of the bar graph, and the smallest are toward the bottom. Both the data map and bar graph represent only one column from one data set at a time. Different instances of visualization are used to compare different variables. The flowchart of the rendering process is shown in Figure 4. The initial rendering of the visualization is demonstrated in Figure 5.

The JavaScript element events mouseover and mourseout are used to synchronize the data enhancement among all the display graphs. When the mouse is on a display SVG element, which is identified by the state ID, all the graph elements corresponding to this ID will receive a signal about the selection, and their color is set to mulberry, which is demonstrated in Figure 6. When the mouse is moved out of the selected element, a signal is sent to all these elements about color changes, and all the corresponding elements will recover their original colors. This event-based scheme used in our platform will dynamically synchronize the state enhancement among all the displayed graphs in real-time. In the screenshot of Figure 5 the state Oklahoma with ID OK is selected and automatically enhanced among all the visualization results on the web browser.

Rendering data progress over time

As shown in Figure 1, the top figure (a) maps the total number of people fully vaccinated per 100. The mouse can show a tooltip of the numerical information with golden color enhancement on the corresponding boundary. The middle (b) is the progress of the boosters per 100 over time, with vaccinations distributed per 100 as the x-axis and people vaccinated per 100 as the y-axis. The bottom one (c) is the progress of deaths per 10,000 over time, with recovered per then 1000 as the x-axis and confirmed per 10,000 as the y-axis. The following items describe the algorithm for displaying the multi-parameter and multi-source data described in Figure 1.

a) Filter data set to ensure all data loaded are valid and have meaningful information. Load $us . geojson$ data for rendering a United States map with 52 states and saved to an array $stateMap$ .

b) Use $d 3 . timeParse$ to load a series of data over 2 years, acquiring data of the percentage of total boosters, total vaccinations, people vaccinated, fully vaccinated, and save the data to an array called $covidData$ .

c) A series of health data about coronavirus over 2 years are loaded using $d 3 . timeParse ()$ , acquiring data of total people confirmed of coronavirus disease $α_{i}$ , deaths due to the disease $γ_{i}$ , the number of people recovered $ϵ_{i}$ , and case fatality ratio $μ_{i}$ (i = 1, 2, …, 52 that represent the 52 states in the United States). Also, the total population of each state $λ_{i}$ of the United States is loaded using an enhanced for loop. Next, use the three loaded data sets to calculate the recovered per 10,000, confirmed per 10,000, and death per 10,000, and then save the data to an array called $vaccineData$ .

d) As shown in Figure 1(a), implement function $vaccinMap ()$ to render and map full vaccination rates per 100 $θ_{i}$ ((i = 1, 2, …, 52) by state using arrays $stateMap$ and $vaccineData$ as input arguments. The array $stateMap$ matches each state with its corresponding vaccination data using $d . properties . name$ comparison and loading. The low color with $rgb (105, 255, 255)$ and high color $rgb (10, 65, 65)$ , are employed to map the data range of the percentage of full vaccination in each state, and tooltip is utilized for showing numerical information.

e) Develop function $bubbleChartVacc ()$ to display the dynamic relationship of people vaccinated $ν_{i}$ , vaccinations distributed $π_{i}$ , and boosters used $ϕ_{i}$ . The array $vaccineData$ is used as an input argument. First, using a filter to select all the nonempty valid data, sort them in ascending order using time as the parameter, and set the display region $Φ_{k}$ (k = 1, 2, …, 5 to represent the five regions of the United States) and the corresponding drawing area. Next, set the x and y axis with the corresponding title. $d 3 . scaleOrdinal ()$ is used to group states into one of five regions and map to the corresponding color. When users move the time slider, the system will acquire the users’ selected time and use $selectedDate . getTime ()$ to get the corresponding time. Next, a function $update ()$ is called to update data to newly selected time points and refresh all the rendered content vaccinations distributed $π_{i}$ , people vaccinated $V_{i}$ , and boosters $ϕ_{i}$ . The data boosters per 100 $ϕ_{i}$ are used as the circle radius with mapping function $ to create a square-root-based scale. Furthermore, a tooltip is designed to display selected information. The result is illustrated in Figures 1(b) and 2.

f) Design function $bubbleChartCovid ()$ to visualize the connections of coronavirus disease confirmed, recovered, and death due to coronavirus infection. The array $vaccineData$ and the total population of each state $Ω_{i}$ (i = 1, 2, …, 52) are used as input argument. The data filtering, sorting, and axis creation are the same as step (e). The Figures 1(c) and 2 show the running result of this function. The element $λ_{j}$ ( $j$ = 1, 2, 3) of the array $vaccineData$ is processed using formula $ψ_{i}^{j} = Ω_{i} / λ_{j} \times 10000$ , where $j = 1$ (total people recovered), 2 (total people confirmed), and 3 (total deaths). When users change the time slider, the new date will be selected, and the $selectedDate . getTime ()$ will be used to match the time as $update ()$ function is called to renew all the rendering results to show the newly updated results.

Results and analysis

Multiple data sets and their associated variables and parameters are visualized dynamically with quantitative numbers to investigate the relationships between vaccination, coronavirus disease cases, and the death rate. Moreover, the progress of vaccination, booster shots, and the percentages of disease, recovery, and death are displayed and compared over time. The initial visualization web pages are illustrated in Figures 1 and 5.

Visual evaluation results demonstrate the benefits of our tool in providing a comprehensive and effective visual representation of important patterns over the past 2 years. This helps researchers and medical professionals gain a better understanding of the coronavirus, thereby enabling them to combat the disease more effectively. Moreover, in the next step of this research, healthcare professionals will be interviewed using our data visual analysis and explanation platform to gather their comments and feedback.

Information enhancement

The developed software platform introduces an interactive synchronization function to enhance feedback using events and signal schemes. When users move the mouse to hover on any bar or state in one of the six graphs, which corresponds to one state, all the other five graphs will enhance to corresponding state information with mulberry color $rgb (224, 75, 153)$ so the users can easily see and analyze the information in all these graphs. The quantitative information is also displayed in the graphs and charts so that users can visualize and synchronize multi-source data sets in one web browser, and the enhancement can be dynamically updated using the mouse.

The Figure 6 demonstrates this function using the state Texas as an example, from which we can see that state Taxes is colored with mulberry in (a), and the age group of this state is also enhanced in (b). The corresponding vaccination and booster information and the coronavirus case and death rates are colored in (c) and (d) in real-time, respectively. At the same time, users can also intuitively detect the enhanced quantitative information of the vaccinations by age group, illustrated in (e) and (f) for showing the percentage of full-vaccinated residents and the percentage of residents with a booster, respectively.

Data mapping

The quantitative information visualized in Figure 5(b) to (d) can be mapped to the United States map (a) using color mapping with $d 3 . scaleLinear ()$ , which is a linear scales mapping a continuous, quantitative input domain to a continuous output range using a linear transformation. The low color yellow $rgb (255, 255, 120)$ and the high color blue $rgb (0, 102, 235)$ are used as range, and the corresponding domain is the minimum and maximum of the visualized data.

We developed a multidimensional data visualization platform that uses a map of the United States to represent the percentage of vaccination, coronavirus case rates, and death rates due to coronavirus disease. Figure 7 shows five screenshots corresponding to the five selection choices in the dropdown menu depicted in Figure 8 for the year 2022. We collected data and created corresponding visualizations spanning from 2020 to 2022. The rendering categories include “At Least One Dose,”“Fully Vaccinated,”“Booster,”“Case Rate,” and “Death Rate,” with the mapping results illustrated in Figure 7(a) to (e), respectively. The mapping domain is determined by the corresponding dataset’s minimum and maximum values for the data column. The mapped range is from yellow to blue, meaning the ranges of the scale bar on the right in each of the five maps differ. For example, the range is from 60 to 95 in the “At Least One Dose” graph, while the coronavirus death rate graph ranges from 0.5 to 1.4. This allows for easy detection of color differences across states on the map.

Figure 7.

Screenshots of multidimensional data visualization on a United States map representing the percentage of vaccination, coronavirus case rate, and death rate due to coronavirus disease: (a) At least one dose, (b) Fully vaccinated, (c) Booster, (d) Coronavirus case rate, and (e) Death rate. The screenshots are depicted in Figure 7.

Figure 8.

Selection of multi-parameter options from the dropdown menu (top left) and corresponding screenshots generated for the United States map representations in Figure 7, depicting: (a) At Least One Dose, (b) Fully Vaccinated, (c) Booster, (d) Case Rate, and (e) Death Rate in the year 2022.

From Figure 7(a) to (d), we can see that when a state has a higher at least one dose rate, this state usually has a higher fully vaccinated and booster rate, and vice versa. When comparing (c) - (d) with (e) and (f), we can detect that when a state has a higher vaccination rate, this state usually has a relatively lower case rate and corresponding lower death rate due to coronavirus disease. Moreover, we can see that the death rate due to coronavirus disease is low, which is <1.4%. We can conclude that coronavirus disease has relatively strong transmission capability, that is, it can be around 40% in the total population, but the mortality risk is relatively low.

We can also extract information that the vaccination is region-based, for example, the northeast region has a relatively higher vaccination and booster rate, which can achieve around 90%–95%, and this region also has a relatively lower coronavirus disease case rate that is <25%. On the other hand, the south region, especially Louisiana, Mississippi, Alabama, Tennessee, Kentucky, and West Virginia, has a relatively lower vaccination rate of around 60% fully vaccinated, and this region has a relatively higher case rate that is around 32%–37%.

Multi-source bar chart

The quantitative information on vaccination rate, including at least one dose, fully vaccinated rate, and booster, can be displayed in 1 bar chart in our system, described in Figure 5(c). We use the merge method to design this visualization, during which the largest data set, that is, at least one dose, is displayed first in the bar chart in gray color, then the second largest data set, that is, fully vaccinated, is visualized in the same bar chart in cyan color, so in each state, the light gray color is covered to the cyan color. Next, we display the smallest data set, that is, booster rate, in blue. The quantitative numbers in each state of the three categories are shown in each bar to analyze and compare this information between the states.

A similar method is used to visualize the quantitative information of case rate and death rate in Figure 5(d). We can detect the relationship between the vaccination rate and case rate using these 2 bar charts. For example, in the states Maine, Maryland, and Vermont, the fully vaccinated rates are 82%, 78%, and 84%, respectively, and the corresponding case rates are 22.5%, 21.3%, and 23,1%, respectively. On the contrary, in the states of Tennessee, Arkansas, and Louisiana, the fully vaccinated rates are 56%, 52%, and 55%, respectively, and the corresponding case rates are 34.3%, 31.2%, and 32.5%, respectively. We can see that a higher vaccination rate can get a lower coronavirus disease rate, which can prove the effect of the vaccination.

However, there are some exceptional cases, such as in the states of Rhode Island and Massachusetts, where the fully vaccinated rates are 86% and 82%, but the corresponding case rates are 40.4% and 30.5%. The death rate is 0.9% and 1.1%, respectively, the case rate is relatively high even with a high vaccination rate, the death rate in both states is among the average, so we also need to consider other factors that can affect the case rate. Figure 6(c) demonstrates Texas state enhanced in vaccination rate, while Figure 6(d) shows the case and death rate with the same state colored.

Age group consideration

Usually, the coronavirus disease rate has connections with the percentage of age groups in the population of a state, so we show the age group of all 52 states in the United States. The percentage of six age groups displayed in Figure 5(b), that is, 0–18, 19–25, 26–34, 35–54, 55–64, and 65 and over. However, we did not detect a direct relationship between the groups in a state with its coronavirus disease rate. Therefore, we also display the vaccination and booster rates in the age groups of all 52 states, which are illustrated in Figure 5(e) and (f), respectively.

The connected scatterplot is used with a bar chart to show the vaccination and booster rates in various age groups. The bar chart is used to connect different age groups in a state and to synchronize information enhancement, that is, when the mouse is moved over the bar of a state, the state’s visual information in all other graphs is enhanced simultaneously in mulberry color, which has been demonstrated in Figure 6 using Texas as an example. As demonstrated by Figure 5(e), for the percentage of fully vaccinated residents, four age groups are visualized, that is, 5–11, 12–17, 18–64, and 65 and up, while for the percentage of residents with a booster category, three age groups are visualized, that is, under 18, 18–64, and 65 and up, which is illustrated in Figure 5(f). From both (e) and (f), we can see that for both fully vaccinated and booster rates, the age group 65 and up has a significantly higher rate than the other age groups, which can achieve above 84% and 51% in all states. In contrast, the corresponding rates for the age group 18–64 are 67% and 18%, respectively. The age group under 18 is much lower than the first two.

From the vaccination rate in the age group analysis, we can conclude that even the senior residents have relatively higher chances of getting coronavirus disease. Still, this age group has a much higher vaccination and booster rate, so the coronavirus disease case rate is not higher than other age groups. We can also know that there is no direct connection between the percentage of age group and the state’s coronavirus disease rate. Figure 6(b) shows part of the vaccination rate with Texas state enhanced with mulberry color, and the corresponding partial figures of the vaccination by the group are described in Figure 6(b) and (f) with the same state selected.

Multi-parameter evaluation over time

As demonstrated in Figure 1, the final information of the percentage of total people fully vaccinated is visualized in (a), multi-parameter vaccination data is dynamically displayed to show the progress over time from 08/27/2021 to 05/10/2023, which is illustrated in (b). The coronavirus disease case number and recovery information are visualized with a time range of more than 1 year using multiple parameters. All the visual information can be updated in real-time and demonstrated in (c). The data in (b) and (c) are grouped into one of five regions with individual colors, and the time slider is designed for navigating the time to show the updated information, so the displayed data will be rendered dynamically.

The dynamic visualization result of the progress of the percentage of vaccination, people vaccinated, and boosters used in four-time points is described in Figure 2, from which we can see that at the beginning of the data rendering, the percentage of boosters is small, the District of Columbia has the largest percentage 14.02% in getting a booster dose of vaccine, and the state West Virginia has the lowest booster dose percentage of 0.07%. The state with the lowest percentage of vaccination distributed is Wyoming, with 100.41%, and the state with the lowest percentage of people vaccinated is Idaho, with 43.85%. The initial rendering result is illustrated in Figure 2(a). As shown in (b), the states of the largest and smallest percentage of getting a booster dose are the same as (a) but increase by 1.6 times. Also, we can see that most states have similar percentages of vaccine booster dose usage, and the states in the Northeast region have relatively high vaccination rates in both distribution and usage. The booster rate difference among the states in the region of the Southeast is the largest, which is in the range from 1.23% to 19.25%. When the time point changes to October 2022, the state of Vermont has the largest booster dose rate of 84.29%, which is illustrated in (d). All the states have increased the vaccination and vaccine booster dose rate, but the speed is different. The Northeast region keeps the highest average of all the three parameters of vaccinations distributed, people vaccinated, and booster dose usage. The overall web-based result of vaccination is illustrated in Figure 1(b).

The dynamic progress of coronavirus disease confirmed, recovered, and death is illustrated in Figure 2. The visualized information of this pandemic disease at the beginning time of 11/10/2020 is shown in (a). New Jersey and New York are the two states with the highest rates of death per 10,000, with 18.09 and 17.48, respectively. However, the confirmed and recovered number per 10,000 in these two states is relatively low, with 301.55 and 42.18, 278.22, and 41.82, respectively. The state of North Dako has the highest confirmed and recovered rate. Also, we can see that in most states, the relationship between the confirmed and recovered rates is close to linear. Most states with high death rates are in the region of the Northeast. As selecting the time point at the end of November 2022, the three parameters of the confirmed, recovered, and death in all the states have increased, but the increasing speed in the first two parameters is faster than the third one. The states in the Midwest region have the largest range in the confirmed and recovered ratios, described in (b).

When we select the next time point of 01/09/2021, as illustrated in Figure 2(c), the relationship between the rate of confirmed cases and recovered from coronavirus is almost linear for most states. However, the four states in the Northeast, that is, New Jersey, New York, Rhode Island, and Connecticut, have relatively low recovery and high death rates. At the same time, there are four states, that is, Vermont, Hawaii, Maine, and Oregon, that have much lower death rates even though they have low recovery rates, and the difference is around nine times when comparing Vermont with New Jersey. We think the isolation and precautions differences among these states need to be further investigated to explain the reason. When we select the end time point using the slider, as shown in (d), the four northeast states, including New Jersey, have relatively high death rates, and the corresponding recovery rate changes little. We can also see that Arizona and Kansas have relatively low recovery rates but high death rates. Also, the relationship between confirmed and recovered rates is almost linear in most states. The two states of North and South Dakota have the highest confirmation and recovery rate. At the same time, Vermont and Maine have the lowest death rates of 3.04 and 4.83 per 10,000, respectively, and both belong to the northeast region.

Here, we have visualized and analyzed the vaccination, dose distribution, booster dose usage, and death rate due to the pandemic. However, other factors such as isolation, wearing a mask when levels of coronavirus disease infections are higher, getting tested and staying home if a person is sick, getting tested if required by the workplace, and wearing a mask to protect the vulnerable should also be investigated and analyzed to explain all the rendering results logically. The difference in the states on the implementation of quarantine measures, application of social distancing measures at schools and workplaces, and lockdown of communities also should be measured and evaluated to understand further the visualized difference among all the states in the United States.

Conclusion

In this research, we have developed multiple innovative algorithms and a software platform for interactive, multi-parameter, high-dimensional pandemic data visualization and temporal synchronization. These advancements offer significant advantages over traditional methods for processing and displaying multidimensional data. We have introduced data rendering and visual analytics tools to display and evaluate multi-parameter epidemiological data, helping users understand pandemic transmission patterns and their interactions with vaccination from multiple perspectives. The developed software platform enables epidemiologists and decision-makers to uncover hidden information about the infectious disease caused by the SARS-CoV-2 virus, such as pandemic behavior and transmission trends, and to correlate these findings with results obtained through machine learning algorithms.⁴⁸ While our focus has been on exploring and analyzing coronavirus disease, vaccination, and demographic data, these innovative algorithms have broader applications in general data visualization, inspection, and uncovering hidden insights.

This research aims to develop dynamic visualizations in various dimensions that consider coronavirus disease and vaccination data. We have developed new algorithms and a web-based software platform to address the lack of publications on connecting and synchronizing pandemic data acquired from various sources on one web page to compare and analyze multiple parameters. Our platform dynamically displays multi-source pandemic data and extracts information such as coronavirus disease rate, number of people vaccinated, vaccinations distributed, percentage of mortality, vaccination distributions among various age groups, case and fatality ratio, booster usage, and their relationships with virus transmission rates over the past 2 years.

Using SVG (Scalable Vector Graphics) elements with D3 (Data-Driven Documents) as JavaScript extensions, we have designed a novel representation-transparent approach for web-based data visualization. Our platform can interactively adjust time points and map multi-source information with various parameters. Using HTML select elements, the choropleth map, bar chart, colored bubble plot, and connected scatterplot render multi-parameter pandemic data on a web page. The platform includes a time slider and tooltip elements to display extracted vital information over time, allowing multi-dimensional variables to be examined without needing to change the code and reload the visualization. Once the graphs and charts are generated, various collective instances of the visualization can be compared to identify hidden patterns from the multi-parameter pandemic data visually. When the system is used on a large scale, we will also consider sustainable and green computing technologies, such as virtualization and cloud computing.

There are some limitations in this research study, such as the data being collected only in the range of <3 years and at the state level. Based on the limited data this paper uses, reaching comprehensive conclusions in the long run will be challenging. Additionally, some information is missing in the data collected, necessitating the use of filters to delete these incomplete records to maintain the integrity of the data items for visualization and analysis.

In the next phase of this research, we will design deep learning algorithms for information extraction and analysis. We also plan to collect more comprehensive data spanning over 3 years and at finer levels of granularity, such as the county level, while incorporating additional parameters. This includes gathering data on socio-economic factors to formulate and test causal hypotheses, linking these factors with vaccination progress and pandemic trends. Additionally, we will interview healthcare professionals to gather their feedback on using our algorithms and how our platform can enhance their understanding of the pandemic. This feedback will help us refine the platform to develop more effective solutions for controlling coronavirus transmission and better meet real-world medical requirements.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Illinois State University, including the Research Grant, the Travel Funding of the School of Information Technology, the University Research Grant, the Publication Incentive Program of the College of Arts and Sciences, and the University Publication Support Program.

ORCID iD

Qi Zhang

References

Onyeaka

Anumudu

Al-Sharify

, et al. Covid-19 pandemic: a review of the global lockdown and its far-reaching effects. Sci Prog 2021; 104(2):368504211019854. DOI: 10.1177/00368504211019854.

Chen

Shi

, et al. Visual data analysis and simulation prediction for COVID-19. Int J Educ Excell 2020; 6(1): 95–114.

Allen

Almukhtar

Aufrichtig

, et al. Coronavirus in the U.S.: Latest Map and Case Count. https://www.nytimes.com/interactive/2021/us/covid-cases.html (2022).

Ribeiro

Nisi

Prandi

, et al. A data visualization interactive exploration of human mobility data during the covid-19 outbreak: a case study. In: 2020 IEEE symposium on computers and communications (ISCC), Rennes, France. pp. 1–6. DOI: 10.1109/ISCC50000.2020.9219552.

Excler

Privor-Dumm

Kim

JH.

Supply and delivery of vaccines for global health. Curr Opin Immunol 2021; 71: 13–20.

Fauci

Lane

Redfield

. Covid-19 - navigating the uncharted. New Engl J Med 2020; 382(13): 1268–1269.

Maxwell

Shreedhar

Dauga

, et al. Fair, ethical, and coordinated data sharing for covid-19 response: a scoping review and cross-sectional survey of covid-19 data sharing platforms and registries. Lancet Digit Heal 2023; 5(10): e712–e736.

Meuschke

Garrison

Smit

, et al. Narrative medical visualization to communicate disease data. Comput Graph 2022; 107: 144–157.

Sadiku

Shadare

Musa

, et al. Data visualization. Int J Eng Res Adv Technol 2016; 12(4): 2454–6135.

10.

Kandiah

Iheaku

Farrque

, et al.; CROSS Collaborative. COVID-19 vaccine hesitancy among health care workers in four health care systems in Atlanta. Open Forum Infect Dis 2022; 9(8): ofac224

11.

Jayanti

Yulyantari

. Interactive data visualization: open data of pandemic disease. In: 2021 3rd international conference on cybernetics and intelligent system (ICORIS), Makasar, Indonesia. pp. 1–5. DOI: 10.1109/ICORIS52787.2021.9649569.

12.

Leung

Kaufmann

Wen

, et al. Revealing covid-19 data by data mining and visualization. In: Barolli

Chen

Miwa

(eds) Advances in Intelligent Networking and collaborative systems. INCoS 2021. Lecture Notes in Networks and Systems. Cham: Springer International Publishing, 2022, vol. 312, pp.70–83. DOI: 10.1007/978-3-030-84910-8_8.

13.

Leung

Chen

Hoi

, et al. Big data visualization and visual analytics of covid-19 data. In: 2020 24th International Conference Information Visualisation (IV), Melbourne, Australia. pp. 415–420. DOI: 10.1109/IV51561.2020.00073.

14.

Qin

Luo

Tang

, et al. Making data visualization more efficient and effective: a survey. VLDB J 2020; 29(1): 93–117. DOI: 10.1007/s00778-019-00588-3.

15.

Bernasconi

Grandi

A conceptual model for geo-online exploratory data visualization: the case of the covid-19 pandemic. Information 2021; 12(2): 69.

16.

Zhang

Sun

Gaggiano

, et al. Visualization design practices in a crisis: behind the scenes with covid-19 dashboard creators. IEEE Trans Vis Comput Graph 2022; 29(1): 1–11.

17.

Padilla

Fygenson

Castro

, et al. Multiple forecast visualizations (mfvs): trade-offs in trust and performance in multiple covid-19 forecast visualizations. IEEE Trans Vis Comput Graph 2023; 29(1): 12–22.

18.

Latif

Usman

Manzoor

, et al. Leveraging data science to combat covid-19: a comprehensive review. IEEE Trans Artif Intell 2020; 1(1): 85–103.

19.

Sharma

Bodempudi

Reehl

. Real-time data visualization to enhance situational awareness of covid pandemic. In: 2020 International Conference on Computational Science and Computational Intelligence (CSCI), Las Vegas, NV, USA. pp. 352–357. DOI: 10.1109/CSCI51800.2020.00066.

20.

Lipeng

Xuedong

Jianqin

, et al. Design and development of a visualization system for covid-19 simulation based on webgis. In: 2020 International Conference on Public Health and Data Science (ICPHDS), Guangzhou, China. pp. 278–282. DOI: 10.1109/ICPHDS51617.2020.00061.

21.

John

Shaiba

. Data visualization as a tool for decision making: Analysis based on covid-19 data. In: 2021 International Conference on Decision Aid Sciences and Application (DASA), Sakheer, Bahrain. pp. 870–874. DOI: 10.1109/DASA53625.2021.9682345.

22.

Rimal

Gochhait

Bisht

Data interpretation and visualization of covid-19 cases using r programming. Inform Med Unlocked 2021; 26: 100705.

23.

Maciejewski

Livengood

Rudolph

, et al. A pandemic influenza modeling and visualization tool. J Vis Lang Comput 2011; 22(4): 268–278.

24.

Akhtar

Tabassum

Perwej

, et al. Data analytics and visualization using Tableau utilitarian for COVID-19 (Coronavirus). Glob J Eng Tech Adv 2020; 3(2): 28–50.

25.

Tao

Yang

Wang

, et al. A survey on data-driven covid-19 and future pandemic management. ACM Comput Surv 2023; 55(7): 1–36.

26.

Engledowl

Weiland

Data (mis)representation and covid-19: leveraging misleading data visualizations for developing statistical literacy across grades 6–16. J Stat Data Sci Educ 2021; 29(2): 160–164.

27.

Banerjee

Paul

Srikanth

, et al. Causal connections between socioeconomic disparities and covid-19 in the USA. Sci Rep 2022; 12(1): 15827.

28.

Kong

Tekwa

Gignoux-Wolfsohn

SA.

Social, economic, and environmental factors influencing the basic reproduction number of covid-19 across countries. PLoS One 2021; 16(6): e0252373.

29.

Raphael

Schneider

Introduction: the socioeconomic impacts of covid-19. RSF 2023; 9(3): 1–30.

30.

Hrin

Emmerich

, et al. Development and validation of a covid-19 vaccine hesitancy scale for adults in the united states. Vaccine 2022; 40(40): 5764–5768.

31.

Covid-19 Vaccines. Vaccines * SS on Widening perspectives on the BAFF family Ligands & receptors. Curr Opin Immunol 2021; 71: 111–116. DOI: 10.1016/j.coi.2021.07.003.

32.

Fan

Fay Siu

Rao

, et al. The accessibility of data visualizations on the web for screen reader users: practices and experiences during covid-19. ACM Trans Access Comput 2023; 16(1): 1–29.

33.

Theparod

Kreabkhontho

Teparos

Booster dose vaccination and dynamics of covid-19 pandemic in the fifth wave: an efficient and simple mathematical model for disease progression. Vaccines 2023; 11(3): 589.

34.

Atherton

“Missing/unspecified”: demographic data visualization during the COVID-19 pandemic. J Bus Tech Commun 2021; 35(1): 80–87.

35.

Westmeier

Paniskaki

Karaköse

, et al. Impaired cytotoxic CD8⁺ T cell response in elderly COVID-19 patients. mBio 2020; 11(5): e02243.

36.

Battle

Heer

Characterizing exploratory visual analysis: a literature review and evaluation of analytic provenance in tableau. Comput Graph Forum 2019; 38(3): 145–159. https://doi.org/10.1111/cgf.13678.

37.

Matlab. R2023b Release Highlights - MATLAB and Simulink, howpublished. https://www.mathworks.com (accessed 12 October 2023).

38.

DOMO. Put data to work and help everyone multiply their impact, howpublished. https://www.domo.com (accessed 12 October 2023).

39.

Denice

Choi

Haan

, et al. Visualizing the geographic and demographic distribution of COVID-19. Socius 2020; 6: 2378023120948715.

40.

Comba

JLD

. Data visualization for the understanding of covid-19. Comput Sci Eng 2020; 22(6): 81–86.

41.

Afifah

Sudaryatno . Effectiveness of classification method and color symbol scheme on choropleth map of population density in special region of Yogyakarta. ISPRS Int Arch Photogramm Remote Sens Spat Inf Sci 2019; XLII-4/W16: 55–61.

42.

Node.js. Version: 18.13.0. OpenJS Foundation and Node.js contributors, 2023. URL https://nodejs.org/en/blog/release/v18.13.0.

43.

Express.: Fast, unopinionated, minimalist web framework for node.js. OpenJS Foundation, StrongLoop/IBM, 2017. https://expressjs.com/.

44.

The coronavirus pandemic: see how vaccinations are going in your county and state. The New York Times, 2022. https://www.nytimes.com/interactive/2020/us/covid-19-vaccine-doses.html.

45.

Demographics and the economy indicators (Population). Population distribution by age. State Health Facts, 2021. https://www.kff.org/state-category/demographicsand-the-economy/population/.

46.

Covid data tracker. Maps of covid-19 vaccinations by age and sex over time. Centers for Disease Control and Prevention (CDC), 2022. https://covid.cdc.gov/covid-data-tracker/#datatracker-home.

47.

USA COVID-19 Vaccinations. Kaggle, 2023. https://www.kaggle.com/datasets/paultimothymooney/usa-covid19-vaccinations.

48.

Alballa

Al-Turaiki

Machine learning approaches in covid-19 diagnosis, mortality, and severity risk prediction: a review. Inform Med Unlocked 2021; 24: 100564.

Multidimensional data visualization and synchronization for revealing hidden pandemic information

Abstract

Keywords

Introduction

Motivation and contributions

Materials and methods

Platform and tools used

Data preparation and processing

Data visualization and synchronization

Rendering data progress over time

Results and analysis

Information enhancement

Data mapping

Multi-source bar chart

Age group consideration

Multi-parameter evaluation over time

Conclusion

Footnotes

Funding

ORCID iD

References