Application of a seeded hybrid genetic algorithm for user interface design

  • Authors:
  • Nicholas Hardman;John Colombi;David Jacques;Raymond Hill;Janet Miller

  • Affiliations:
  • Air Force Institute of Technology, Wright-Patterson, Ohio;Air Force Institute of Technology, Wright-Patterson, Ohio;Air Force Institute of Technology, Wright-Patterson, Ohio;Air Force Institute of Technology, Wright-Patterson, Ohio;711th Human Performance Wing, Wright-Patterson, Ohio

  • Venue:
  • SMC'09 Proceedings of the 2009 IEEE international conference on Systems, Man and Cybernetics
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Studies have established that computer user interface (UI) design is a primary contributor to people's experiences with modern technology; however, current UI development remains more art than quantifiable science. In this paper, we study the use of search algorithms to predict optimal display layouts early in system design. This has the potential to greatly reduce the cost and improve the quality of UI development. Specifically, we demonstrate a hybrid genetic algorithm and pattern search optimization process that makes use of human performance modeling to quantify known design principles. We show how this approach can be tailored by capturing contextual factors in order to properly seed and tune the genetic algorithm. Finally, we demonstrate the ability of this process to discover superior layouts as compared to manual qualitative methods.