Digital Reminder is a system which manages a person's view images based on gaze detection. A real-world-oriented database that consists of one's view is constructed by one's implicit behavior of watching objects. Digital Reminder provides a user with WYSIWYR- (what-you-see-is-what-you-retrieve-)based interaction for view image retrieval. WYSIWYR is a new framework in information retrieval. In retrieving image data in a database, we generally enter a query by specifying one or more image attributes such as color or shape with a keyboard and/or a pointing device. Unlike former image retrieval, WYSIWYR framework identifies the user's current view of watching as a query condition, and the user's behavior of watching an object is regarded as a trigger for the retrieval. Since the context of watching, that is, the type of visual information and gaze duration, is also recognized as a view frame attributes (VFA), a more appropriate result is presented to the user. The result of performance evaluation showed that the view images similar to one's current view of gazing were ranked higher in the result of retrieval, compared with the retrieval without VFAs.