Even though human memory is critical in our daily lives, it eventually becomes ambiguous and difficult to recall. In this research, we propose a text-based lifelog approach with attributes that capture useful features for memory recall as well as three lifelog ranking functions (an event ranking and two “remember” rankings) using features that can be calculated from a user’s lifelog. We evaluated these ranking functions and developed a prototype system based on our approach. We found that a combination of the following is useful: event duration, place, person, social media post, and photo, especially place and photo.