Organizational decision-making increasingly relies on Decision Support System (DSS) tools. During the past 30 years, research on machine learning has enabled these DSS tools to become progressively more intelligent [1]. In particular, machine learning algorithms have enabled DSS to learn and to be responsive to changing decision-making environments. In this paper, we aim to expand on how machine learning can be applied to a dynamic decision environment where we can improve our understanding of customer satisfaction based on online product or service reviews. Customer review websites, such as TripAdvisor for travel, Yelp and Urban Spoon for restaurants, and Patagonia, Lands’ End and Epinions for product reviews have become commonplace. These sites allow customers to both read and provide reviews of a product or service. The customer reviews, which are often openly available on the web, contain a wealth of information usable by management, competitors, investors, and other stakeholders to discover the customer concerns which drive overall customer satisfaction for a particular service or product. For simplicity, we use the terms ”overall satisfaction” and ”overall rating” interchangeably. The term ”aspect” denotes a real or perceived component or feature of a product or a service that matters to the customer. Traditionally, customer satisfaction questionnaires would be used to determine the significant components, or aspects, of overall customer satisfaction [2]. However, questionnaires are expensive or may not be available. In some cases, public agencies are even prohibited by law from collecting satisfaction questionnaires from customers. In situations such as this, the only alternative may be to ana lyze publicly available free-form text comments, for example, from sources such as TripAdvisor. Even in cases where customer satisfaction surveys are collected, analysis of free-form text comments available on the Internet can provide important insights usable in isolation or in conjunction with traditional closed-form customer satisfaction questionnaires.