Recommendations in Product Personalization: An Extended Example of Pipelined Filter Combination
Volker Renneberg, Uwe M. Borghoff · Zenodo (CERN European Organization for Nuclear Research) · 2003
Traditionally, customers spend most of their money for pre-configured "off the-shelf" products. Simple product variations are offered, fine-tuning by the customer before the purchase is generally not intended. Why pay for functionality that is not wanted? Why spend money for products that are so rigid and do not satisfy a hundred percent? To remedy these issues, we discuss a possible solution based on recommendations. In our approach, traditional filtering techniques known from recommender systems guide the customer through a sophisticated product personalization process using pipelined filter combination. This process distinguishes between expert customers and novices. An extended example proofs the rich capabilities of the proposed solution.