You are working for a consulting firm that performs data mining for a large, multi-national credit card issuer named ABCard that has several card types and customer databases in various countries, some of which have been recently acquired. Because of these recent acquisitions, some of the source databases are still in the native countries’ languages and currencies. Your market basket analysis of ABCard’s customers and their purchases shows some strong associations that are extremely profitable to the corporation. However, these strong and profitable associations also reinforce gender stereotypes that might be seen in a negative light. Before presenting your data mining analysis to ABCard’s marketing team, your manager asks you to go through the knowledge discovery and data mining steps to make sure everything is correct. Required:
Explain what you would check at each step in the process (1. data cleaning, 2. data integration, 3. data transformation, 4. selection, 5. mining and 6. pattern evaluation) to ensure that your conclusions are correct.