In this study, we seek to automate the categorization of attributes from online reviews based on the Kano model. We present a framework which relates customer feedback at an attribute specific level with overall feedback at the product or service level and therefore infers the nature of the attribute’s influence on customer experience. We validate this framework on a large-scale data set of a popular portal that covers restaurant ratings globally. Our data covers approximately 65,000 reviews across ten cities, ten different cuisines, and four attributes. Our analysis results in various location-cuisine specific insights regarding the attributes. At a high-level, our analyses show that the food quality as an attribute tends to behave like a necessity driver when the cuisine is native, and as a luxury driver when it is non-native. We present three case studies which illustrate that managerial insights at location-cuisine specific level can be inferred on each of the aspects, and would, therefore, allow for a restaurant-specific prioritization of the aspects. © 2018, Curran Associates Inc. All rights reserved.