Stubkjær: Real Estate and the Ontology of a Multi-Disciplinary Domain

Real Estate and theits Ontology ofwithin a Multi-Disciplinary Domain

Prof. Erik Stubkjær

Dept. of Development and Planning, Aalborg University,

Fibigerstræde 11, DK 9220 Aalborg East

Denmark

, fax +45 9815 6541

Extended abstract submitted for

*** Geographical Domain and Geographical Information Systems ***

EuroConference on Ontology and Epistemology for Spatial Data Standards

La Londe-les Maures, France, 22-27 September 2000

Abstract

Among spatially extended objects, units of real estate make a distinct category. The units' complex relationships with other phenomena are of economic, legal, political, and spatial nature, and even technological with respect, for instance, to its representation on cadastral maps. A multi-disciplinary project has been proposed to establish a coherent knowledge base in this field. The project focuses onreal propertythe transactions of real property within European countries, and it addresses the ontology of real estate, or terrain object, and of its grounded boundaries and idealised limits.
Real estate is legally defined, and the related property and land-use rights count among the basic institutions of most societies. It is suggested, that the ontology of this domain is formalised in dialogue with the ontology development of the corresponding university teaching. Furthermore, it is contended that the ontology development has to be based on a philosopher's notion of ontology, which differs from the ontological conception of database entities by the knowledge engineer.
Keywords: Spatial objects, cadastre, institutions, multi-disciplinary research, ontology, real estate, real property rights, real property transactions.

1. The diverse notions of ontology

The philosopher Barry Smith points to the fact that the term ontology is used differently by philosophers, and inif compared with the fields of information processing and information technology respectively (1999). This observation is important when synergies are expected from relating research in both ontology with research in andknowledge representation.

Smith develops on the different understandings by the philosopher-ontologist and the ontological engineer, and characterizes their main concerns. In search for a method to bridge the different understandings, he refers to the attempts within biology to construct ontologiesthat apply to of the term 'gene' and similar biological species. The methods include the constructionelaboration of common vocabularies, and the formulation of appropriate translation rules between the diverse nomenclatures of the different branches of biology. Biologists cooperate with software (ontology) engineers, as well as with philosophers in this endeavor. This approach motivates similar efforts in other domains, see section 2 below.

Smith assumes that "(e)very scientific field will .. have its own preferred ontology, defined by the field's vocabulary and by the canonical formulations of its theories" (Smith, 1999: 1). This position may hold for the disciplines that were established before, say, the 1950s, but not for the many university fields that have been establishedsprung/blossomed due to practical trends, societal needs, technological releases and the vast increase of enrollment at university. Only a segment of university teachers prioritize the development of a formal ontology of their domainnew field. It is assumed that they face the same problems as ontology engineers, and may apply the same methods.

2. Ontology development and multi-disciplinary university studies

The remainder of the paper explores the potential of a strategy of ontology development that benefits from the ontology developmentprocess of ontologies in multi-disciplinary university studies. The author uses his own domain: cadastral studies, as an example.

Multi-disciplinary studies are faced with constant changes within their target section of society, as well as changes due to new research outcomes from the established university disciplines and professional scopes they draw upon. A proactive strategy for handling those changes is to establish a formal ontology for the domain concerned as a crucial step for its theorization. As a further benefit, the development of rich and rigorous knowledge structures of the domain will, potentially, allow for accumulating generally valid domain knowledge, putting away theoretical fallacies and insignificant buzz-words.

The development of formal ontologies as an independent research strategy is faced with the same challenge of constant changes , as are thein any multi-disciplinary studies. If theour ontology development is based on the philosopher's basic ontological units like man, society, matter, reality, representation, and language, possibly theour research outcome will be of universal scope. However, research projects may be limited in time and resources, more that multi-disciplinary teaching and related research programmes. Thus, one may wonder whether - in a 5 years perspective - ontology development that is generated mainly by research funds and domain knowledge from industry or licenced professions, developsgrows faster than ontology development in the context of the corresponding university teaching. These considerations suggest,that the ontology of the Real Estate domain is formalised inby a dialogue with the formal ontology development of the corresponding university teaching. Consequently, the development of the multi-disciplinary domain of cadastral studies is described, as it provides one context for the development of a formal ontology of Real Estate and the related spatial phenomena.

3. Real Estate, an entity within the cadastral problem domain

In Europe, the cadastre developed in the context of centralization of administration and the issuing of tax ordinances,fromsince about 1700 (Encyclopaedia, 1930). Cadastral concerns at university level branched from geodesy and land surveying. In the USA North America, the branching can be evidenced through the specification of research needs (Stubkjær & Ferland, 2000). The university teaching of law, i.e.: Cadastral law, property lawrights, and planning legislationregulations, took place in a national setting, and still does. However, from the 1970s a concern for an international scope manifests itself, e.g. in terms of the Statement on the Cadastre, Cadastre 2014: A Vision for a Future Cadastral System, and the recent Badthurst Declaration on Land Administration for Sustainable Development, bothwhich were prepared by the "«Fédération Internationale des Géomètres»" (International Federation of Surveyors) (FIG, 1995; 1998, 1999).

Those normative statements offrom the surveyors' profession has been accompanied by research, e.g. from the point of view of formal modelling (Frank, 1996), benchmarking (Steudler, Williamson, Kaufmann & Grant, 1997), or with a view to chart the interrelated technical, legal, and organisational aspects (Zevenbergen, 1998) and the technological conversion of a cadastral map to digital database (ex: CERTU, 1996). Stubkjær surveys research in information systems development, and research within geographical information science (GISc), with a view to establish a theoretical basis for cadastral studies (1999),and thus presents a view of the cadastral problem domain (2000). This view was drafted with reference to the Soft Systems Methodology (SSM) in its early version (Checkland, 1981). The view is graphically rendered through Figure 1 above.

Stubkjær & Ferland discuss the design of research within the multi-disciplinary cadastral domain (2000). SSM suggests a 'political system analysis', but SSM does not provide the concepts and methods of Political science. Knowledge of these concepts and methods are needed in order to develop a valid ontology of the domain. At least in principle, the same applies regarding the knowledge of the other established disciplines. However, within a single research project it is hardly feasible to monitor and apply recent research from several disciplines. This leads to the following suggestion for a research strategy: The elements of the cadastral problem domain (cf. Figure 1) are described with reference to main works of the relevant disciplines through an international standing project. Individual research projects, e.g. of the PhD-type, can then draw upon this 'cadastral knowledge base' and - within the scope of the project - update the base by investigating the most recent research.

A multi-disciplinary, European project has been proposed as a COST action research to establish a coherent knowledge base in this field. The project focuses on real propertythe transactions of real property within various European countries, and it addresses the ontology of real estate, or terrain object, and of its grounded boundaries (Stubkjær, 1999cost), its ideal or abstract limits or its indefinite confines.

4. Approaching the ontology of Real Estate

The proposal for COST Action: Modelling Real Property Transactions, draws - among others - on the fiat-bona fide dichotomy of Barry Smith (1995), among others. To give examples: abona fide boundary like a stream may be used for the definition of a fiat boundary, e.g. the boundary of a jurisdiction. Fiat boundaries may be established as mere abstract lines traced on a plan for the division of land, to structure its settlement. However as time passes on, the owner occupies and uses the terrain until the confines with the neighbour's one and then, letting marks of their reciprocal possession, the lines become visible and permanent on in the terrain as the buildings, fences and roads are established that is, they become bona fide. The fiat-bona fide distinction can make a basis for comparisons of the boundary setting practices of different countries, as well as for investigations into topological problems that extend to several dimensions, what is called mereo-topology (Smith & Varzi, 2000). More, such distinctions can handle neighbouring entities of different types of consistency and limits, as an individual private property parcel-lot and an open-field commons or a domanial forest.

As supportive technologies develop, property boundaries arebecame located by means of a national, geodetic reference system. However, a cadastral location needs more than the absolute position with reference to the geodetic referencecoordinates systems, as owners of the estate do not understand the language of geodetic referencing, and. The neighbour relations among the estate units cannot be represented by absolute coordinates (Laurini & Thompson, 1992). The neighbour relations appear in the landscape accordingly to the structure of the settlement, on the analogue cadastral map with parcel identifiers, and through the corresponding data structures. Also, the units need to be described relatively to place names, especially road names, as this is the location meansfashion that is used by the ordinary citizen and owner.

Using the scales of measurement of Stevens (1946) in the simplifiedan adapted version where the ratio scale is ignored, we have (Stubkjær, 1992; Stubkjær & Ferland, 2000):

Nominal (verbal) / Place names; Cadastral identifiers; Address codes
Ordinal (graphical, topological) / Neighbour relations; House numbering sequences
Metric (numerical) / Coordinates of boundaries, road centerline, etc.;
'Metes and bounds'.
Ratio (numerical, graphical) / Regular/ proportional land divisions in lots along baselines; Sub-numbering of lots identifiers relative to parcel partition

Stubkjær suggest that the unit of ownership is categorized as a jurisdiction, which is distinguished from other classes of socio-economic units: place name, district, and region (Frank et al, forthcoming, GISDATA series 9).

Summing up, the concepts presented may assist the development of a formal ontology that would be is independent of the rule sets and practices of a specific European country, but that can adequately encompasses most of the specific cases.

1 of 1 .