Abstract/Summary

Purpose – Automated valuation models have been in use at least for the last 50 years in both academia and practice, while automated valuation recently re-emerged as very important with the rise of digital infrastructure. The current state of the art, therefore, justifies the dual contributions of this paper: organising existing knowledge and providing a new framework. Design/methodology/approach – This paper provides much-needed analysis and synthesis of the accumulated body of knowledge by proposing an updated classification of automated valuation approaches based on two criteria, and a taxonomy adapted to new trends. The latter requires a paradigm shift from models to automated valuation systems. Both classification and taxonomy arose after literature review. Findings – This paper provides a framework for an explicit context under which automated valuation is carried out. To do so, authors propose a definition of automation valuation systems; contextualise the differences among theories, approaches, methods, models and systems present in automated valuation and introduce a classification of automated valuation approaches and a non-hierarchical taxonomy of automated valuation systems. Research limitations/implications – Perhaps, a systematic literature review process instead of a selective list of 100 references could additionally validate the proposed classification and taxonomy. Practical implications – The new framework, underlying various dimensions of the automated valuation process, can help practitioners surpass judging models based purely on their predictive accuracy. Also, the automated valuation system is a more generic term that can better accommodate future research coming from a multitude of disciplines, more diverse business areas and enlarged variety of practical users. Originality/value – This is the first paper that develops a taxonomy of automated valuation systems.