Mining spatial association rules in census data

In this paper we propose a method for the discovery of spatial association rules, that is, association rules involving spatial relations among (spatial) objects. The method is based on a multi-relational data mining approach and takes advantage of the representation and reasoning techniques developed in the field of inductive logic programming (ILP). In particular, the expressive power of predicate logic is profitably used to represent spatial relations and background knowledge (such as spatial hierarchies and rules for spatial qualitative reasoning) in a very elegant, natural way. The integration of computational logics with efficient spatial database indexing and querying procedures permits applications that cannot be tackled by traditional statistical techniques in spatial data analysis. The proposed method has been implemented in the ILP system SPADA (spatial pattern discovery algorithm). We report the preliminary results of the application of SPADA to Stockport census data. 1. Background and motivation Censuses make a huge variety of general statistical information on society available to both researchers and the general public. Population and economic census information is of great value in planning public services (education, funds allocation, public transportation), as well as in private businesses (locating new factories, shopping malls or banks, as well as marketing particular products). The application of data mining techniques to census data, and more generally, to official data, has great potential in supporting good public policy and in underpinning the effective functioning of a democratic society [29]. Nevertheless, it is not straightforward and requires challenging methodological research, which is still in the initial stages. As an illustrative example of some research issues, let us consider the census data table reported in Figure 1, where each row represents an enumeration district (ED), the smallest areal unit for which census data are published in UK (). () National statistics institutes (NSIs) make a great effort to collect census data, but they are not the only organisations that analyse them: data analysis is often done by different inst itutes. By law, NSIs are prohibited from releasing individual responses to any other government agency or to any individual or business enterprise, so data are summarised for reasons of privacy before being distributed to external agencies and D. Malerba, F. Esposito, F. A. Lisi and A. Appice  Mining spatial association rules in census data 20 The data analyst might be interested in finding some kind of dependence between the active population and the percentage of cars per household. A dependence can be expressed as an association rule, that is an implication of the form