Abstract
This short paper summarizes challenges and opportunities of applying machine learning methods to Web design problems, and describes how structured prediction, deep learning, and probabilistic program induction can enable useful interactions for designers. We intend for these techniques to foster new work in data-driven Web design.