Uncovering patterns of disadvantage in student socioeconomic data using k-means clustering and association rule mining

This study employs a combination of k-means clustering and association rule mining to analyze socioeconomic data from 507 students at a community school in Udon Thani Province, Thailand, highlighting the pronounced economic inequalities impacting educational opportunities. By integrating these data mining techniques, the research categorizes students into four income-based clusters and identifies critical socioeconomic attributes influencing educational access and quality. The application of k-means clustering revealed four distinct economic categories within the student population, with the lowest income cluster significantly lacking in basic amenities such as air conditioning and agricultural vehicles, which are essential for their living and educational environments. Association rule mining was then applied to identify household attributes linked to the ‘very low’ income classification. Fourteen association rules were discovered, with support values ranging from 0.43 to 0.44, confidence levels between 0.81 and 0.82, and lift values consistently at 1.1. Key findings reveal that the absence of basic amenities such as air conditioning, agricultural vehicles, computers, and televisions are strongly associated with the lowest socioeconomic status. For instance, the rule {AirCondition=N, TV=N} => {Cluster=Very Low} had a support of 0.44, confidence of 0.81, and lift of 1.1, indicating households lacking these items have an 81% probability of being classified in the ‘very low’ income group. Notably, the presence of electricity did not correlate directly with higher income clusters, indicating that basic infrastructure access alone does not mitigate the broader socioeconomic challenges. These findings underscore the complex interplay between income levels and access to educational and basic living resources. The study’s implications are profound, advocating for targeted educational policies and interventions that address the nuanced needs of economically disadvantaged students, thereby enhancing their educational outcomes and socioeconomic mobility. Future research directions include broadening the dataset to multiple schools across different regions and conducting longitudinal studies to reveal long-term effects of socioeconomic factors on education.