In the modern Web, users are accessing their favourite Web applications from any place, at any time and with any device. In this setting, they expect the application to user-tailor and personalize content access upon their particular needs. Exhibiting some kind of user- and context-dependency is thus crucial in Web Engineering. In this research, we focus on separating the adaptation engineering process from regular Web engineering by applying aspect-oriented techniques. We show how semantic information and metadata associated with the content can be exploited in our aspect-oriented approach. Furthermore, the approach allows the use of global (structural) properties of the Web application in adaptation specification. We thus obtain several advantages, which are demonstrated in this paper: to control adaptation (specification) separate from (regular) Web Engineering concerns in a richer, more consistent, robust and flexible way.