heterogeneity which adds more complexity to the problem of federating heterogeneous GIS. [Laurini, R., 
1994] listed some discrepancies that could exist in different GIS databases: spatial representation, spatial 
level, projection and coordinated systems, spatio-temporal difference are some examples. Following 
Molenaar’s classification of the types of geometric information we can arrive to a more structured list of 
possible geometric discrepancies that could exist between heterogeneous GISs. There are three aspects of 
geometry: topology, size and shape, and position and orientation. Also there are three levels of topology: 
connectivity of geometric elements, connectivity of objects, and connectivity of geometric elements to 
objects. Thus, it is not suggested to develop a protocol for object identification based on its geometry. 
In the above section it is demonstrated that developing a mechanism for object identification and exchange, 
in a heterogeneous federated GIS, based on syntax or geometry is complex and may result an unstable 
system. It is rather more practical and intuitive to develop a mechanism for object exchange at the semantic 
level because geometry and syntax are prone to change in a database during life time of an object while its 
semantic is more stable. 
4. Syntax And Semantics in GIS Theory 
Current applications of data sharing allow users to exchange their data on the syntactic level, the ODBC 
solution solve the syntactic problem. There are several programs for raster to vector conversion. Several 
research papers proposed mapping between several data models with some discrepancies in their 
hierarchies. However, because the data stored in the database are showing only the syntactic specifications 
it enforces users to base their queries on it. Following the idea of the seven layers of networking protocols, 
data on the syntactic level could be made on the semantic level by building semantic layers on top of such 
syntax. In this case object identification in a federation will be based on its semantics. The system will 
automatically resolve the syntactic and schematic discrepancies ■ 
between the data requester and the data provider. 
Relation Between 
Contexts 
The GIS theory introduced by [Molenaar M., 1993 (a) (b)] 
[Molenaar M., 1994] and [Molenaar M. et al., 1994 (a) (b)] 
properly defined the building blocks of spatial objects. The 
building blocks are forming the GIS syntax Figure 2. At the lowest 
level of the syntactic definition we find the classic data structures, 
i.e., field and object based approaches. The GIS theory formalizes 
the topologie relationships amongst objects, uncertainty aspects, 
and the handling of geometry and topology of fuzzy objects. 
Finally the theory introduces a consistent framework for object 
hierarchies, i.e., generalization, aggregation, etc. 
Relation Between 
Views 
-Semantics 
Relation Between 
Extentions 
Relation across 
Hierarchies 
Object Hierarchies 
Uncertainty and 
Fuzzy Relations 
Topology 
-Syntax & 
Schemata 
Field and Object 
Based Structures 
Figure 2 Syntactic and Semantic 
Definition 
Figure 2 shows the semantics that have to be built onto the syntax. 
There is no consensus on the meaning of syntax and semantics. 
Object class hierarchies, which we consider as syntactic problem, 
could be viewed as a semantic problem by another research group. 
However, we consider the object class, i.e., intension, hierarchies 
as a syntactic problem while the functional relationship between them, within and across hierarchies, as 
semantic problem. This is similar to the functional relationship between objects [Date C.J., 1995]. The 
relationship between hierarchies is the second semantic level. The relation between contexts, i.e., different