Over the years, a lot of course material has been developed to explain to undergraduate students the fundamentals of XML, and schema languages such as DTD and XML-Schema. Typically, the syntax of these languages is discussed and examples are given. How to find a schema for some XML content is often not covered by the material. As a result, students have problems to start with modeling a complex schema, many of their inferred XML schemas are too liberal, and some are even incorrect. In this paper we present a systematic approach for modeling XML content models based on rewriting regular expressions. A small-scale experiment has demonstrated that the quality of the models is improved, and that the approach helps students to begin modeling XML content.