Representing web data in a machine-understandable format is an important task for the next generation of the web. Most solutions reported to date rely on ontologies. However, the use of ontologies is associated with many problems. Moreover, representing web data in a machine-understandable format is not sufficient to respond to all user queries. This paper proposes an approach for representing dynamic web page content retrieved from an underlying database in the concept description language (CDL). This is a semantic format that does not depend on ontologies. However, CDL describes the semantic structure of web content based on a set of predefined concepts and semantic relations. Moreover, this work proposes the addition of extra knowledge to the semantic level of the DB schema to improve the query answering process. A prototype of the proposed approach was implemented to demonstrate its feasibility, and a simple query engine was developed to demonstrate the effectiveness of adding extra knowledge to the DB semantic schema.