Maximal Marginal Relevance-Based Recommendation for Product Customisation

Customised product design is attracting increasing attention. However, consumers can be overwhelmed by the variety of products. To confront this challenge, paper presents a two-step recommendation approach for customised First, an adaptive specification process captures customer requirements in accelerated manner presenting most informative attribute to specify. Then, maximal marginal relevance-based set presented, based on customer’s partial specifications. This ensures broad coverage customers’ needs considering not only relevance each their but also redundancy set.