A Rule Mining Approach to Emotional Design in Mass Customization1

In the current manufacturing environment of intense competition, mass customization becomes a crucial capability for a company to survive. Acquiring and identifying customer needs is an important aspect of mass customization. Emotional terms are often included in the description of customer needs and customer evaluations. At the same time, products that consider emotion responses and evoke positive emotions tend to attract customers. Thus, emotional design has been proposed to incorporate customer emotional needs into design elements to deliver customer emotional satisfaction. The main challenge of emotional design originates from difficulties in mapping customers’ subjective emotions to perceptual design elements. This paper proposed an association rule mining approach to provide emotional design decision support where the basic requirements can be identified. This approach applies rule mining on customer evaluations and product specifications. Three main steps— data preprocessing, association rule mining, and rule evaluation—are involved in this approach. Data preprocessing extracts critical emotional terms and perceptual design elements from data sources. Data provided by similar customers are grouped. For each customer group, association rule mining is applied on the emotional terms and associated design elements to find out implicit relationships between them. Support level, confidence level and freshness of supporting data of the rules are used to evaluate and rank the applicability of the rules. The proposed approach is illustrated by a case study on emotional design for a series of mobile phones.