Unveiling the Hidden Truth of Repurchase Intention: A Textual Approach Using Aspect-Based Deep Learning

  • Zusheng TAN (Speaker)

Activity: Talks or PresentationsOther Invited Talks or Presentations


The advancements in web technologies have fueled the proliferation of the voice of the customer (VOC) on the Internet; such VOC is often used to predict the repurchase intention of individual customers. Nowadays, customer repurchase intention has been a major topic in both academia and industry. On the one hand, researchers want to understand how different VOC (e.g., product/service description and evaluation) affect repurchase intention. On the other hand, organizations want to utilize these VOC to offer the best possible product/service to their customers to retain them and create positive repurchase intentions. Regarding this, our study proposed a new annotation scheme to construct an open multi-aspects VOC dataset, namely the Restaurant Corpus for Customer Repurchase Intention (RCCRI). Our dataset can be used to explore different repurchase-related aspects in VOC like customer attitude and behavior, usage experience, and usage comparison (e.g., customers often compare products/services when writing reviews). Our dataset differentiates us from prior studies, which often encode repurchase intention as a single value (0: negative; 1: positive) and focuses on only one or few aspects (e.g., food, services, atmosphere). Additionally, we propose a new approach, namely the aspect-based repurchase intention prediction (ABRP), which captures individual linguistic factors in VOC that contribute to customer repurchase intention.
Period3 May 2022
Event titlePostgraduate Seminar Series
Event typePublic Lecture