Angelina Tzacheva ¹, Sanchari Chatterjee ², Rajia Shareen Shaik ³ and Shiva Sai Praneeth Chakinala ³, ¹ Westcliff University, USA, ² University of North Carolina at Charlotte, USA, ³ Walmart Inc, Bentonville, USA

In the modern world of data, data mining focuses on techniques to extract surprising, engaging, and previously unknown patterns of knowledge from massive datasets. Extracting this data is beneficial in multiple domains. This paper explores Action Rules as a framework for extracting actionable insights from large-scale data in education and business. We introduce a Modified Hybrid Action Rule Mining approach with Information Granules and Meta-Actions. We assess the Cost and Feasibility of the discovered Action Rules. Our proposed method enhances scalability, efficiency, and interpretability through Big Data analytics. Experiments on student survey datasets and Net Promoter Score (NPS) business datasets demonstrate improved performance in transitioning emotions (e.g., Sadness to Joy, Detractor to Promoter). Our Results show that cost and feasibility of each Meta Action empower users to make informed, goal-oriented decisions.

Action Rules, Data Mining, Cost and Feasibility.