SUMMARY
The increase in poverty rates caused by the COVID-19 pandemic requires immediate attention from policymakers. Each province in Indonesia has unique characteristics of poverty, and as a result, each province's response to COVID-19's impact on poverty is unique. As a result, a provincial cluster analysis based on the similarity of poverty characteristics is necessary to identify provinces that require increased vigilance. The purpose of this study is to cluster Indonesian provinces according to their similarity in terms of poverty impact before and during COVID-19. The impact of poverty prior to and during COVID-19 is quantified by comparing 2021 (during COVID-19) to 2019 (before COVID-19). We discovered that the COVID-19 has a significant impact on poverty. Hybrid SOM-Kmeans with three clusters is the optimal method for producing the smallest Davies-Bouldin Index. COVID-19 has high, moderate, and low impact on poverty, respectively. Cluster 1 is a cluster with a significant impact on poverty in a province where tourism is the primary industry. Due to sluggish tourism, the community's purchasing power is diminished, thereby increasing poverty. Cluster 3, namely Papua, has a low impact due to its primary sector characteristics in the mining sector.