Search

Cadastral data are maintained through formally defined procedures which need to provide security and consistency. Databases and transactional model enable consistency but lack flexibility in modelling business processes and support for heterogenous IT environments (web services, various programs). Transactional workflow management systems (WFMS) provide flexibility and can provide consistency of data. Land administration domain model (LADM) provides an excellent basis for modelling static component of land administration systems, but doesn’t provide elements to model dynamic component, i.e. processes. In this paper we define conceptual model of a dynamic component of land administration system. We use the WFMS concept with integrated transactional support. Data model enables storing elements of Petri nets and it is divided to generic and extended. Generic model ensures consistency of processes on object level and is applicable on cadastral data generally. Extended model ensures consistency by spatially defining affected area of the process and it is used to model processes on cadastral parcels spatially represented by polygons. Modelling of processes and transactional support is achieved with Petri nets. Workflow elements enable ensuring consistency of a process in a pessimistic or optimistic manner. Pessimistic approach ensures consistency by locking objects affected by the process and optimistic approach leaves checking of concurrent changes until the very end of the process. Finally, we demonstrate how the devised model copes with a simple example of two separate processes where each wants to split one of two adjacent parcels, in a pessimistic manner.

The primary purpose of a Land Administration System (LAS) is to support the processes of recording and disseminating information about the ownership, value and use of land and its associated resources. An important task of LAS is land valuation for taxation purposes, which supports land management processes. Land valuation for taxation implies mass valuation methods, which use statistical analysis of valuation indicators to assess the best use of many real properties at the same time. However, to make fair and correct assessments, sufficient accurate and up-to-date data for each land parcel together with accompanying buildings and other constructions on it must be available. Some of the required data exists, but they are stored across different land registers and other registers, and are administered by various public authorities. To avoid further data redundancy, mass valuation information system can be based on existing data stored in distributed registers and shared by agreements between various public authorities. Therefore, the crucial part of the LAS is the data model, which needs to be carefully developed to efficiently fulfil the primary purpose of the system ; its development should be based on best practices from countries with similar LAS using well-known standards. This paper analyses the data models in the official Croatian registers and the possibilities of using them in land valuation and taxation extension in the Croatian Land Administration Domain Model (LADM) profile. The analysis identifies possible data, registers and responsible authorities whose data are appropriate to be used as an attribute for determining valuation indicators. The basic unit of such a system is real property, which can be generally seen as land and accompanying buildings or other constructions, or jointly owned individual parts of the property according to the property subdivision plan. Regardless of the valuation method used, each property type valuation indicator can be divided into external factors (e.g. location, zoning regulations) and internal factors (e.g. land parcel area, number of floors) for each property and can be seen in general as real property characteristics (e.g. land parcel area, number of floors). Some of the required data are found to be redundant (e.g. personal data) and some are found to be missing (e.g. building characteristics). For multiple sources of data, the paper considers the most appropriate one, based on the assessment of data completeness and whether the data are up- to-date. Schema matching is used to relate valuation attributes to appropriate LADM classes. Schema matching is a set of techniques used for comparing the schema of different data models, searching for similar or equivalent elements. The manual schema matching approach proved to be appropriate for the relatively small extent of the covered Croatian LADM data. The created UML class diagrams are used to show the level of compliance of available data with LADM valuation and taxation extension. This work demonstrates that a detailed review of valuation attributes is crucial in the development process for a mass valuation system because the whole system greatly depends on the availability and quality of the chosen attributes. An appropriate mass valuation system based on these premises can improve the usability of existing LAS, which could be used not only to provide fair and efficient land taxation, but also as a tool to make urban and rural spatial planning more sustainable.