Measuring Relevance in Support of Design Reuse from Archives of Building Product Models

Abstract

This paper investigates the use of text analysis techniques to measure relevance between objects in product models. This relevance measure is used to support design reuse from an archive of product models from previous projects. We consider several techniques from the field of information retrieval for comparing documents. Next, we explore whether we can improve the performance of the relevance measure by considering contextual building model elements. The paper concludes with an evaluation and discussion of the techniques considered. We conclude that relatively good retrieval results can be achieved with short texts and sparse building models such as those that occur in current industry practice.