During day-to-day activities, a user refers to various data items on a PC, such as documents, images, and audio clips. However, when the number of items in the history increases, locating the data becomes more complicated. We propose a novel system called "OreDesk" to retrieve particular data items from the history. OreDesk calculates the "Active Index" and "Relation Index" values for each data item based on the user operations. Active Index indicates the interest of a user in the data. Relation Index indicates how closely one data item is related to another. By using these indices, data can be retrieved on the basis of the association of user experiences. OreDesk also provides users with facilities such as "Searching Tool for Related Data" and "Timeline Viewer". OreDesk is available at http://oredesk.net/ as a free software.