Constraint driven model using correlation and collaborative filtering for sustainable building

Sustainable building has emerged as an important topic due to the fact that it can significantly reduce the impact of buildings and their operation on the natural environment and more efficiently utilize resources throughout a building´s life-cycle. When compared with a traditional buildingdesign process, integrated project delivery has proven to be more efficient, and is thus gaining wider acceptance for many sustainable building projects. However, managing design and construction from different disciplines is still challenging. Conflicts among constraints are often not identified at the right design stage, which results in multiple iterations of the design process. In this paper, a novel constrain-driven model that enhances design processes through better management of constraints and thus delivers optimal design solutions with higher energy performance is proposed. Multiple Correspondence Analysis was applied to capture the correlations between different items (parameter-value pairs) and classes (constraints). Meanwhile, it integrated Collaborative Filtering methods and Constraint Satisfaction Problem to train and refine the proposed model. Finally, we have applied our model to a synthetic data sets to demonstrate its performance.