Abstract
Understanding how the consumer perceives quality is a key issue in supply chain management. However, as the market structure continues to deepen, traditional evaluation methods using SEVRQUAL are unable identify all issues related to customer quality and unable to supply solutions. The maturation of data mining technology, however, has opened the possibilities of mining customer attribute data on quality problems from unstructured data. Based on the consumer perspective, this research uses an unsupervised machine learning text mining approach and the Recursive Neural Tensor Network to resolve the attribution process for undefined quality problems. It was found that the consumer quality perception system has a typical line-of-sight that can assist consumers quickly capture the logical structure of the quality problem. Although attributions related to quality problems are very scattered, a highly unified view was found to exist within each group, and a strategy to solve the undefined quality problem was agreed through group consensus by 61% of the consumers.
Recommended Citation
Zhu, Qing; Wu, Yiqiong; Li, Yuze; and Zuo, Renxian, "A Text Mining Based Approach for Mining Customer Attribute Data on Undefined Quality Problem" (2018). WHICEB 2018 Proceedings. 64.
https://aisel.aisnet.org/whiceb2018/64