C-RWD: A Computational Responsive Web Design Service

Abstract

With the increasing diversity of device screen sizes used by web users, responsive web design (RWD) has become an essential part of website production. However, due to its complexity, RWD often requires the cooperation of a team of interface designers, developers and data analysts and requires the team to invest much time. Besides, it is unrealistic to create personalized, responsive designs for users using manual RWD methods.

This thesis proposes C-RWD, which is a computational responsive web design service based on the LaaS platform. The design and implementation of C-RWD are introduced in detail, and the functions of C-RWD are evaluated based on example websites. When an initial interface is provided, C-RWD can (1) automatically collect the interaction data of different users in the interface and generate personalized breakpoints and (2) automatically use combinatorial optimization to generate optimized interfaces at these breakpoints based on user's interaction data to improve usability and then (3) automatically integrate the optimized interfaces under different breakpoint widths into one fully responsive website and present it to users.

The contribution of C-RWD is that it proposes a novel framework which can automate the generation of responsive interfaces while optimizing the interface for user's personalization. In practical applications, it can greatly reduce the time and cost of RWD production and generate a personalized and optimized interface for users.