Abstract
Today people are storing increasing amounts of personal information in digital format. While storage of such information is becoming straight forward, retrieval from the vast personal archives that this is creating poses significant challenges. Existing retrieval techniques are good at retrieving from non-personal spaces, such as the World Wide Web. However they are not sufficient for retrieval of items from these new unstructured spaces which contain items that are personal to the individual, and of which the user has personal memories and with which has had previous interaction. We believe that there are new and exciting possibilities for retrieval from personal archives. Memory cues act as triggers for individuals in the remembering process, a better understanding of memory cues will enable us to design new and effective retrieval algorithms and systems for personal archives. Context data, such as time and location, is already proving to play a key part in this special retrieval domain, for example for searching personal photo archives, we believe there are many other rich sources of context that can be exploited for retrieval from personal archives.
Highlights
Vannevar Bush could never have envisaged the impact his 1945 article ‘As We May Think’ [1], in which he presented his Memex vision, would have on modern science
While much attention has been given to the generation of these vast personal archives (Human Digital Memories (HDM)), e.g. [2][3], less attention has been given to retrieval of information from them
MediAssist [7] associates context data such as time, location, people and weather conditions with photos to allow people retrieve based on their memory of photos. While these systems have made the first steps towards the use of context data to allow people retrieve from their personal archives, based on what they remember about items, they are quite limited in that they only capture a subset of the many ways people remember items
Summary
Vannevar Bush could never have envisaged the impact his 1945 article ‘As We May Think’ [1], in which he presented his Memex vision, would have on modern science. HDMs are fundamentally different from traditional content archives for which existing retrieval techniques have been developed, in that: an HDM is typically a combination of many types of media, audio, video, images, and many texts of textual content; there is the potential for a large percentage of noisy data in these archives; many items in the archive may be very similar, repeatedly covering the same topic; a user may not be aware that a particular piece of data was captured, and is available for retrieval; the user may not be able to describe clearly what they are looking for; items may not have formal textual descriptions, meaning that they cannot be retrieved using standard text or meta-tag based retrieval methods; and items may not be joined by inter-item links, meaning link structure could not be utilized in the retrieval process It is this unique combination of attributes of HDMs that motivate this research into creation of retrieval techniques for the personal archive domain.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
