An entity-based social peer-to-peer network for digital life memories

Abstract

People capture the memorable events of their lives in a digital form in order to share them in their social network. A memory for Life system can organize them in a meaningful way for future use. The main challenges needed to overcome the sharing of digital memories are: defining a suitable network topology, searching amongst appropriate peers for the correct data, data privacy, and the sharer's control over their data, even after it has been shared. To share these events, we propose a novel Entity-based social peer-to-peer network that organizes the network according to the memory threads of entities. In this context, an entity could be a person, place, object or interest. Digital memories are organized in the form of memory threads for each entity. We then organize the network to reflect the memory threads in the network. This idea is based on a realistic approach taken from existing online social networks; this allows us to organize the network in a meaningful way to support social activities and produce better search results. Data privacy is achieved by avoiding the use of a single authority collecting personal information. We have also considered the strength of a peer's social ties in order to find contents, which are more likely to be of social interest to a user. Our experimental results, based on the strength of social ties, show the decrease in network overhead, which is achieved by avoiding irrelevant search responses.