2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) An Interactive Platform to Track Global COVID-19 Epidemic Zhenhe Pan * , Dhruv Mehta * , Anubhav Tiwari * , Siddhartha Ireddy * , Zhou Yang † and Fang Jin † * Department of Computer Science, Texas Tech University Email:{zhenpan, Dhruv.C.Mehta, Anubhav.Tiwari,siireddy}@ttu.edu † Department of Statistics, George Washington University Email: zhou yang@gwmail.gwu.edu, fangjin@email.gwu.edu being able to halt the spread of this and future unknown infectious disease outbreak more effectively and efficiently. In this paper, we developed an interactive visualization platform 1 to closely monitor the global COVID-19 situation and the social media & news reports. Besides showing the latest COVID-19 cases numbers, the platform preserves the historical records and enable users to see the COVID-19 spreading trends. In addition, the platform collects and digests the tweets streams and grab people’s top concern, provid- ing users a One-stop experience of the overview COVID- 19 situation. To summarize, our platform has the following contributions: • Developed a comprehensive database of coronavirus cases worldwide, which would tremendously helps re- searchers and policy makers to model the spread of this infectious disease timely. • Provided a Bird’s-eye view of the global COVID-19 dynamics, not only showing the latest confirmed cases numbers, but also allow users to view the COVID-19 trends for each country and state/province. • Developed practical tools to collect information on coro- navirus cases world-wide by mining multiple online plat- forms. • Integrated breaking events and people’s concern flows into the map, providing a full picture of the COVID-19 dynamics. II. SYSTEM FRAMEWORK Figure 1 illustrates the system framework of our platform. The framework is divided into three main components: (a) A front-end component which collects real-time data from multiple websites and social media platforms; (b) A database storage engine that enables real-time update and historical data retrieve; (c) A back-end component that includes multiple algorithms for tweets filtering, concern flow extraction, and breaking news identification. A. Data Flow As shown in Figure 2, our input data has three types: real-time COVID-19 cases for each country/state, coronavirus related tweets discussions, breaking news from multiple web- sites. The real-time COVID-19 case numbers were collected 1 http://worldcovid-19stats.000webhostapp.com/covid.html#trends Abstract—This project built a world-wide database of coron- avirus cases, which helps to model the spread of the coronavirus disease (COVID-19), and to identify policy and social factors that impact the spread of COVID-19. Four essential tasks are implemented: 1) build a comprehensive database of coronavirus cases world-wide; 2) visualize the heatmap of confirmed cases for each country, provide detailed spreading trends for each countries and comparison among countries; 3) collect tweets about COVID-19 in real-time and extract people’s daily concern flow; 4) i ntegrate b reaking n ews s uch as fi rst confirmed/death case in each country. This demo will provide decision-makers with accurate data-driven representations in an easy to understand format that enables them to make more timely and cost-effective preparation and response plans. I. I NTRODUCTION COVID-19 is spreading rapidly and has already affected more than 169 countries, infecting near a million people and causing more tens of thousand deaths around the world (as of April 1, 2020). The collection of coronavirus cases information such as positive diagnoses, recovery, and death, is vital as it provides a foundation for modeling the disease evolution dynamics, allowing us to infer future trends, and analyze the factors that influence the speed at which the virus spreads. This information will be crucially important if we are to develop appropriate policies to manage this pandemic. A number of platforms [1]–[7] developed to predict and provide people an updated information about coronavirus as this outbreak unfolds. These work increased people’s awareness of this disease and advanced the state of the art by developing accu- rate methods and effective tools to collect coronavirus cases from mining online platforms. However, most of the currently available coronavirus databases have some limitations: they either contain only the current numbers of cases and their locations, completely ignoring the historical information, or only cover certain countries/regions, or lack of timely update. Moreover, the pandemic-related data is stored in multiple unconnected resources (websites, news reports, social media posts, and ad-hoc databases) and recorded in many different languages. This lack of inter-database connectivity, along with the associated language barriers, significantly h amper our efforts to deal with this dangerous pandemic. There is thus an urgent need to develop a comprehensive database to identify and track coronavirus cases world-wide, with the ultimate goal IEEE/ACM ASONAM 2020, December 7-10, 2020 © IEEE 2021. This article is free to access and download, along with rights for full text and data mining, re-use and analysis. 948 2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) | 978-1-7281-1056-1/20/$31.00 ©2020 IEEE | DOI: 10.1109/ASONAM49781.2020.9381436