There are two unsolved problems in the field of virtual learning environments: a set of new types of assessment is required for learning management systems (LMSs), and there is a need for a way to assess lifelong adaptive competencies. Proposed solutions to these problems need to preserve the interoperability, reusability, efficiency and abstract modeling already present in LMSs. This paper introduces a competency assessment data model (CADM) being developed as part of adaptive evaluation engine architecture (AEEA). AEEA is designed to solve the above-mentioned problems while fulfilling all quality requirements. The CADM is described with a data centric model using XML for each assessment scenario.