Time is money: dynamic-model-based time series data-mining for correlation analysis of commodity sales (Q2297097)

The authors wish to determine how to transform the transaction records into sales time series that are suitable for determining correlation commodities via data-mining. The authors look to uncover, in the absence of customer information, how to use time series data-mining to find related goods, and then enhance the sale of goods. The authors look to pinpoint the optimal time for product promotion and cross-selling. A dynamic model with time series data-mining, to promote the correlation analysis of commodity sales, has been introduced by the authors. The dynamic model comprises some distance models that are created via a distance function. The commodity sales correlation analysis process involves four steps. First, data transformation converts transaction record data into commodity sales time series data that take into account time factors. Second, commodity sales time series data are used to establish a dynamic model that reflects sales correlations among different commodities. Third, the related algorithm is designed to execute time series data-mining. Finally, affinity propagation clustering and K-nearest neighbors are used to uncover the patterns and knowledge conducive to undertaking the marketing activities that can increase sales volumes.