CROWDSOURCING DATA TO VISUALIZE POTENTIAL HOTSPOTS FOR COVID-19 ACTIVE CASES IN INDONESIA

As the COVID-19 outbreak spread worldwide, multidisciplinary researches on COVID-19 are vastly developed, not merely focusing on the medical sciences like epidemiology and virology. One of the studies that have developed is to understand the spread of the disease. This study aims to assess the contribution of crowdsourcing-based data from social media in understanding locations and the distribution patterns of COVID-19 in Indonesia. In this study, Twitter was used as the main source to retrieve location-based active cases of COVID-19 in Indonesia. We used Netlytic (www.netlytic.org) and Phyton’s script namely GetOldTweets3 to retrieve the relevant online content about COVID-19 cases including audiences’ information such as username, time of publication, and locations from January 2020 to August 2020 when COVID-19 active cases significantly increased in Indonesia. Subsequently, the accuracy of resulted data and visualization maps was assessed by comparing the results with the official data from the Ministry of Health of Indonesia. The results show that the number of active cases and locations are only promising during the early period of the disease spread on March – April 2020, while in the subsequent periods from April to August 2020, the error was continuously exaggerated. Although the accuracy of crowdsourcing data remains a challenge, we argue that crowdsourcing platforms can be a potential data source for an early assessment of the disease spread especially for countries lacking the capital and technical knowledge to build a systematic data structure to monitor the disease spread.