Floss in Cadastre and Land Registration

Building blocks of OSCAR

The registry aspect of the land administration domain requires that each juris-

diction develop their own data model for implementation in a database of the 

 entities and relations that characterize national laws and regulations regarding the 

rights, restrictions and responsibilities that accompany land ownership. The refine-

ment of a generic or standard data model somewhat reduces the effort required 

for this and allows for some integration. However, this still implies fragmentation 

of development effort especially for supporting software. Each country will also 

have to model the domain processes which would typically be implemented in 

software code (such as data entry and management forms). These schema and 

processes will vary due to differences in language, culture, common practice, laws 

and legal definitions etc. Hence, it is unlikely that a single data model can unify all 

the possible variations, or that a single software package can provide all of the 

necessary process requirements. Hence, different nations will have to create their 

own database design and customize their application software around a common 

code base to implement the registry aspects of the land administration domain 

 relative to their own reality and needs.

---

43

Initial research investigating cross-jurisdictional schema variation has shown 

that this problem can be addressed with the implementation of a process or 

“event” model (as opposed to a simplified structural or “state” model which re-

quires  re-engineering each time it is applied in a new and potentially different con-

text). Data schema may also vary due to evolving requirements which become 

more  sophisticated as the land administration infrastructure becomes more ma-

ture. This temporal schema evolution is considered conceptually similar in nature 

to cross- jurisdictional schema variation. Solutions that are coded based on what is 

known at the time will be difficult and costly (in terms of the maturity of develop-

er resources in developing nations) to modify. Importantly, it is expected that 

these systems will require more modification over time than systems developed in 

economically advanced nations where infrastructure is more mature, well-under-

stood and less prone to change. Hence, it is essential to consider the dynamic 

 evolution of schema so that the effort required to evolve a mature land records 

management system is reduced.

The building blocks of the OSCAR project consider both of these types of vari-

ation in the hope of providing a solution that can evolve between countries (and 

therefore provide a foundation for a FLOSS community to develop) and over time 

(without major development effort) as each nation matures in it’s use of the com-

mon functions programmed into the generic code base. The proposed solution 

also  address issues of data integration (which would be difficult if each country de-

veloped its own schema independently) and the lack of human resources by not 

 necessarily requiring fundamental code to be (re-)developed by each jurisdiction. 

Rather, it is expected that functions can be built within and around the conceptual 

model outlined by Hay and Hall (2009). In essence, this approach conceives of an 

“instrument” which signifies a legal change associated with the status of a land 

parcel through a formal registration and land surveying process. Instruments link 

Agents (people such as surveyors and conveyancers), organizations (such as banks 

and government departments), groups (such as land developers and companies) 

to Objects (a Register Object such as a parcel of land, an apportionment of a par-

cel, or a building that exists on a parcel or a unit within an apportionment) via 

Events which implement the temporal aspects of land administration. Together in-

struments, agents, objects and events form workflows (or business processes) and 

capturing the relations between instruments and documents (such as titles, cadas-

tral parcel boundaries, survey plans etc.) as they move through the land adminis-

tration process is the foundation of the OSCAR approach.

Recent research in the area of semantic markup and ontology for the cadastral 

domain argues for the use of semantic web technologies especially in the area of 

data integration (for both applying updates and cross-jurisdictional data sharing). 

This also highlights the inflexibility of the structural or relational model for both 

domain and temporal modeling (Spéry et al., 2001; Schuurman and Leszczynski, 

---

44

2006). Process and temporal aspects of the domain noted above are also the sub-

ject of recent research with specific importance placed on the concurrent defini-

tion of spatial data (in this case parcel boundaries and their physical and title-

based attributes) and associated process models (van der Molen, 2002; Albrecht et 

al., 2008). In addressing these issues, the architecture proposed for OSCAR exter-

nalizes the domain terms and concepts in the form of a domain ontology made up 

of resources that describe and link concepts and terms within the domain. The 

 Resource Description Framework (RDF) 

28

 is used to detail the associations and re-

lationships between data items and their types. This, together with the use of a 

process or event based architecture, allows a highly dynamic data model to be 

used and results in the conceptually straightforward database design and develop-

ment approach described at a high level by Hay and Hall (2009).