Title Mining Association Rules to Determine the Over-spending Behavior Among Low Income Households in Malaysia

Abstract

Overspending behavior in a household can significantly affect the financial burden, debt accumulation, stress, and economic problems. Spending behavior is one of the financial literacy indicators that empowers individuals to make informed financial decisions, budget effectively, and plan for the future. This study proposes an association rules mining approach to investigate the spending behavior among households with income below 40% (B40) in Malaysia. For this purpose, we employ the Apriori algorithm on 2016 and 2019 Malaysia households' income and expenditure survey data obtained from the Department of Statistics Malaysia to discover over-spending items that occurred in household expenditure. The results showed that up to three associated overspending items were discovered based on several support and confidence settings. There are significant changes in spending behavior in the 2016 and 2019 data. Besides food as the main overspending item in 2016 data, other items such as miscellaneous items, restaurants and hotels, and services were overspent in 2019 data. Moreover, three associated items were found only in the 2019 data. This finding benefits the government in improving financial literacy or implementing effective initiatives to improve the nation's living standards.