It has recently become possible for personal computer(PC) users to use large-capacity storage media in business and home situations. As various kinds of tasks are processed with PCs, too many files have been continuously generated. However, most files are not effectively used and we can currently only basically use keyword-based searches. Therefore, when we search for files with keyword searches and several similar files are found, it is impossible to know which file corresponds to which task. If we can recall a context corresponding to individual files and tasks in this situation, we may be able to associate each file with its task, and finally find the target file and its task easily. We will explain the methodology we used for extracting file-operation behaviors and visualizing the behavior of users, which was more efficient than with the conventional keyword-based searching approach, to recall the context and identify the target file and its task.