Data Mining for Revealing Relationship Between Google Community Mobility and Macro-Economic Indicators

Abstract

Google community mobility reports have helped to evaluate the effectiveness of government-imposed movement control among countries. However, the relationship between the mobility data and the characteristic of regions is less reported. This study aims to reveal hidden information from Google community mobility reports and relate them to all 34 Indonesian provinces' macro-economic indicators. This secondary research implements a data mining approach using the CRISP-DM process framework and Knime Analytics Platform. The community mobility data of residence and workplace are collected as a time series covering Feb 16, 2020, to Jan 31, 2021. Macro-economic indicators are collected from the website of the Indonesian national statistics agency. The clustering method has grouped provinces into three based on their mobility. The findings indicate the relationship between mobility fluctuation during the COVID-19 pandemic and macro-economic indicators, namely human development index and labor force participation rate. In the theoretical aspect, this study has been initiating the investigation of community mobility and macroeconomic. Policymakers in dealing with post-pandemic recovery planning might consider the cluster characteristics for better planning.