Translating the clickstream data in Web server log files into useable information about the characteristics of the consumers who visit Web sites has been a major challenge. This book describes two novel contributions to addressing this challenge. The first is the application of top-down grammatical analysis to clickstream data to infer consumer characteristics. Like top-down approaches to the grammatical analysis of language, this approach specifies sequential patterns to look for in the data, as the patterns marketers are interested in (e.g., purchasing) may be rare, and therefore difficult to detect using bottom-up, data-driven methods. The second contribution is the introduction of a new construct, the Web ad schema, which is a visitor''s mental representation of the structure of a Web site. If this representation is incongruent with the actual structure of a site, a consumer will generate a misleading clickstream. This book should be especially useful to Web site marketers interested in understanding who their customers are and optimizing the information they deliver to their customers, and to researchers developing new applications to analyze clickstream data.