A gold mine of product aspects : analyzing Amazon consumer reviews using text mining

• Main Authors: Tay, Elizabeth Ka-Yin, Quek, Ching Yee
• 其他作者: Poong Oh
• 格式: Final Year Project
• 語言: English
• 出版: Nanyang Technological University 2022
• 主題: Social sciences::Communication::Communication theories and models
• 在線閱讀: https://hdl.handle.net/10356/155856

Online shoppers rely on product reviews when making purchase decisions. This is because where listed information in e-commerce contexts is often inadequate, consumer-generated reviews are a ‘gold mine’ of helpful product information for decision-making. Faced with hundreds, possibly thousands of reviews per product listing, what exactly do consumers seek out when reading reviews? Addressing this question benefits both retailers and consumers through offering the former more effective strategies that can help the latter make better informed choices. The present study adopts topic modeling to extract key product aspects in reviews and performs a series of hierarchical regression analyses to examine how the various factors influence perceived review helpfulness. Subcategory topics were common across all four product categories, while topics related to observable product features and subjective product evaluation were only relevant to experience goods and high-involvement goods, respectively. The mixed effects of review, review author, and product listing characteristics, as well as extracted key topics on review helpfulness call for more in-depth investigation into online consumer behavior, particularly their motivations and how they affect product evaluation processes. The major findings of the current study are expected to inform e-commerce platform improvements that benefit both retailers and consumers.