An Association Rule Mining approach to explore the dynamics in plastic recycling business
Abdullah Al Hasan, Quazi Hamidul Bari, Philipp Lorber, Islam M. Rafizul, Jobaer Ahmed Saju, Eckhard Kraft · Cleaner Waste Systems · 2024
Understanding plastic recycling practices is vital for policy intervention. Association Rule Mining (ARM) is a powerful tool for extracting insights from complex data, though it hasn't been used for plastic recycling analysis before. This study aimed to apply ARM to data from the questionnaire survey of Recycling Shop (RS) owners in Khulna City to identify patterns of recycling practices and recommend suitable policies. Key findings revealed that RS owners rely on local and external sources for plastic waste (Rules with support 0.061–0.242, confidence 0.667–0.8, lift > 1), recommending stronger recycling supply chain policies. Significant links between sourcing and impurities (Rules with support 0.061–0.091, confidence 0.5–0.667, lift > 1.2) suggest better quality control. Disposal methods of non-recyclables like burning (Rules with support 0.03–0.212, confidence 0.714–1, lift > 1) suggest policies for non-recyclables. Using ARM in this study offers a novel approach to developing efficient, sustainable waste management strategies in Khulna City. • Understanding recycling practices is vital for enhancing plastic recycling and reducing plastic pollution. • Identified patterns and relationships in plastic recycling practices in Khulna City using data mining techniques. • An innovative application of Association Rule Mining to analyze plastic recycling data. • The findings can help create smarter recycling programs for the environment and people's health. • Data-driven approaches could transform waste management practices globally.