The Online Customer: New Data Mining and Marketing Approaches By ...

that patterns generated from each cluster, while similar to each other within the cluster, are very different from the patterns generated from other clusters. We further develop a modeling framework for building segment-level predictive models based on the pattern-based clustering approach and signature discovery techniques. We evaluate our pattern-based clustering and model-building approach using different experiments involving 90 different datasets generated from the Web. In each experiment, we compare our approach with several segmentation approaches in data mining and marketing.

In the second essay, we develop original analytical models that address issues related to free shipping promotions and Internet shopping behavior. In order to empirically test the hypotheses generated from the models, we use both common marketing approaches and a data mining method.

The contributions of this research are as follows: (1) We develop a novel pattern-based clustering approach and demonstrate the strength of incorporating unique concepts in data mining and data mining techniques to solve marketing problems. (2) We develop a segmentation-based modeling framework based on a pattern-based clustering approach and signature discovery techniques. (3) We develop original analytical models to derive free-shipping policies, and empirically study issues associated with free-shipping promotions for online shopping. (4) By studying two marketing problems using both data mining and marketing methodologies, we provide insights about the value of both research streams in addressing marketing problems.