Simulating and Modeling Dual Market Segmentation Using PSA Framework

Market segmentation refers to the analytical process of dividing a broad market into segments taking into account multiple factors such as consumer needs, interests and tastes; it has been considered one of the most important marketing strategies as it helps a business to identify hidden market trends, define target segments, and design marketing plans. Market segmentation may also be viewed as a computational challenge: Given the massive amount of data describing interactions between consumers and commodities, the task is to partition the set of consumers and commodities into subsets that corresponds to market segments - two consumers are in the same segments when they exhibit a similar purchasing pattern, while two products are in the same segments when they are purchased by a similar group of consumers. In this work, we focus on the definition and simulation of market segments. We employ the Propose-Select-Adjust (PSA) framework, introduced in an earlier work [10], to simulate the forming of market segments. Our approach is distributed and can be applied to large and dynamic market data set. The experimental results suggest that the proposed approach is a promising technique for supporting intelligent market segmentation.