Activity-based Information Retrieval: Technology in Support of Personal Memory

Much information is hard to retrieve because the need to do so was not foreseen at the time the information was stored. This problem appears hard to solve with the aid of computers. However, research in the area of autobiographical memory suggests the use of retrieval cues relating to the kinds of activities in which the user was engaged at the time of storage. These are highly correlated with the time at which the event took place, and thus can be used for indexing, taking advantage of the accurate time-stamps that computers usually apply automatically when information is stored. In the proposed activity-based retrieval system, data on the user’s activity are gathered automatically to support retrieval by, for example, browsing through an ‘autobiography’ of work episodes. The paper discusses some of the difficulties with this approach, and presents the results of three experiments in which autobiographical episodes were reconstructed by means of automatically collected activity data. It concludes with some comments on the technical feasibility and social acceptability of such an approach to information retrieval.