Rule-Based Promotional Segmentation In Higher Education Using Apriori Algoritm

Abstrak

This study aims to identify co-occurrence patterns among higher education promotional media and translate them into rule-based promotional recommendations using association rule mining. An exploratory quantitative design was applied to 118 anonymized promotional engagement transactions from Institut Teknologi dan Bisnis Muhammadiyah Wakatobi. Each transaction consisted of three promotional items from ten media and content categories. Data were processed through item-label standardization, transaction transformation, binary one-hot encoding, frequent itemset generation, and association rule extraction using support, confidence, and lift metrics in Google Colaboratory. The results show that Facebook information was the most frequent promotional item (support = 0.4322), followed by registration form (0.3814), study program video (0.3559), and Instagram campaign (0.3220). The strongest association rule was Social media → Facebook information, with confidence of 0.4839 and lift of 1.1195, indicating positive association beyond random co-occurrence. The findings were translated into a six-segment rule-based promotional framework. This study contributes to educational data mining by providing an interpretable and replicable approach for supporting data-driven promotional planning in higher education

Kata kunci

• Apriori algorithm
• association rule mining
• educational data mining
• higher education promotion
• promotional segmentation