The potential application of data mining techniques in the extraction of information from property data sets is discussed. Particular interest is focused upon neural networks in the valuation of residential property with an evaluation of their ability to predict. Model testing infers a wide variation in the range of outputs with best results for stratified market subsets, using postal code as a locational delimiter. The paper questions whether predicted outcomes are within the range of valuation acceptability and examines issues relating to potential biasing and repeatability of results.

Case-based reasoning is an artificial intelligence technique which utilizes past experience to solve current problems and in this respect it mirrors the process involved in real estate appraisal. This paper investigates its application as a computer-assisted valuation tool to the specific domain of retail rent determination. As property appraisal is goal orientated, it is essential that the most appropriate examples of previous rent determinations are selected. 1n exploring a case-based reasoning approach to the retail real estate domain five models are built namely; pure inductive, inductive (Q-model), inductive (prototype), inductive (Q-model and prototype) and nearest neighbour.

Rents at review are commonly assessed by making comparisons with properties similar to the subject premises under assessment. Subjective adjustments are made where the location, physical and lease term characteristics of subject and comparable properties differ. The methods employed in such determinations have severe limitations. This paper reviews the various computer-assisted techniques employed in the valuation domain and selects a case-based reasoning approach (CBR) for the determination of retail rents at review. In exploring the application of CBR to retail property valuation five separate approaches are developed, namely; nearest neighbour, pure inductive, inductive (Q Model), inductive (Prototype) and inductive (Prototype and Q Model). The validation of the models demonstrates that CBR is an effective approach to comparable selection in the determination of retail rents.