Schemas can be defined on different levels of abstraction:
In HALE we define the schema mapping on the conceptual level, between classes/types and properties. For the complete transformation process however, information about the data format is needed. The approach used in HALE can deal with both kinds of schemas, abstracting a conceptual view for logical schemas, while using the information from the logical schema for the transformation process. For conceptual schemas, no transformation can be applied, but the mapping can be exported or reused for logical schemas that were derived from the conceptual schema.
Logical schemas are usually more readily available compared to conceptual schemas, as they describe how the actual data is stored, and are either defined explicitly or can even be derived from the existing data structure.
A schema defines a set of types or classes. Each type has a set of properties. A property either has an associated property type or is a group of properties. Properties with a property type have a structure according to that type's definition, they may have a value and properties of themselves.
The image on the right displays a schema as represented in the Schema Explorer view. It allows exploring a schema and inspecting the schema elements. The top level elements in the tree are types defined in the schema, their children are the type's properties, their children being the properties of the properties etc.
See the Schema elements reference for a detailed explanation on how schema elements are represented.