The paper "Deriving the Pricing Power of Product Features by Mining Consumer Reviews" by Nikolay Archak, Anindya Ghose, and Panagiotis G. Ipeirotis explores the impact of consumer reviews on product sales and preferences. The authors argue that traditional methods, which often rely on numeric variables representing review valence and volume, fail to capture the multifaceted nature of product reviews. They propose a text mining technique to decompose textual reviews into segments, evaluating different product features and their associated opinions. The study uses a 15-month dataset from Amazon, covering digital cameras, camcorders, and PDAs, to test their approach. The results demonstrate how consumers' relative preferences for different product features can be learned from these reviews. The paper also discusses two experimental techniques to address data sparsity and omitted variables, and highlights the managerial relevance of the findings for extracting actionable business intelligence from user-generated content.The paper "Deriving the Pricing Power of Product Features by Mining Consumer Reviews" by Nikolay Archak, Anindya Ghose, and Panagiotis G. Ipeirotis explores the impact of consumer reviews on product sales and preferences. The authors argue that traditional methods, which often rely on numeric variables representing review valence and volume, fail to capture the multifaceted nature of product reviews. They propose a text mining technique to decompose textual reviews into segments, evaluating different product features and their associated opinions. The study uses a 15-month dataset from Amazon, covering digital cameras, camcorders, and PDAs, to test their approach. The results demonstrate how consumers' relative preferences for different product features can be learned from these reviews. The paper also discusses two experimental techniques to address data sparsity and omitted variables, and highlights the managerial relevance of the findings for extracting actionable business intelligence from user-generated content.