Observation of sales trends by mining emerging patterns in dynamic markets

One of the regular tasks for managers is to formulate various marketing strategies to confront competitive and changing market environments. Therefore, identifying product sales trends in order to evolve a better strategy is an important issue. It is also an issue that disturbs managers because the preferences of customers often change and capturing them is not an effortless process. In the data mining field, emerging patterns (EPs) are defined as itemsets of which supports increase significantly from one dataset to another. EPs, therefore, can capture emerging trends in timestamped databases or useful contrasts between data classes. In this study, we integrate the concepts of product life cycle and emerging patterns to define four emerging patterns, referred to here as Growth patterns, Rapid-Rise patterns, Decline patterns, and Rapid-Sink patterns, so as to identify interesting sales trends. In addition, we propose a new approach to discover these four emerging patterns from two real-life datasets. Experimental results show that the proposed approach can help managers identify these four types of patterns and interesting trends.