A project restarting support system using the historical log of a user's window usage

This paper proposes a project restarting support system using the historical log of window usage on a computer desktop. The system allows users to simultaneously reopen files associated with a main file by automatically detecting important windows and associated windows from a log. By using our system, users can manage windows in a project without having to manually group windows. We conduct an experiment to evaluate the system. The results indicate that the system can accurately identify as important windows the main files a user used for projects by clustering keyboard operation times on windows; it can also detect files associated with the main file without detection omissions by clustering window z values and visible representation ratios.