The proliferation of interactivity between Web content producers and consumers underscores the development of the Internet in recent years. In particular, customer reviews posted on the Web have grown significantly. Because customers represent the primary stakeholder group of a company, understanding customers’ concerns expressed in these reviews could help marketers and business analysts to identify market trends and to provide better products and services. However, the large volume of textual reviews written in informal language makes it difficult to understand customers’ concerns. This paper describes an integrated approach to summarizing customer reviews. The approach consists of the steps of sentence extraction, aspect identification, sentiment classification, and review summarization. We report preliminary results of using our approach to summarize product reviews extracted from Amazon.com. Our work augments existing work by considering nonstandard input and by incorporating linguistic resources and clustering in automatic summarization.