Watching Through the Web: Building Personal Activity and Context-Aware Interfaces using Web Activity Streams

This paper proposes the use of the increasing numbers of Webbased user activity and personal information sources to enable the creation of more personal, adaptive and activity-sensitive information tools. We describe our initial steps at investigating this idea, including challenges surrounding integrating information from heterogeneous web data sources. This paper contributes an implementation of an in-browser framework called PRUNE that solves the data integration problem by building an internal world model consisting of an entity database and event chronology. We demonstrate PRUNE's aggregation and simple modeling facilities in an application called Notes that Float which automatically-learns associations between notes taken and a user's other activities to enable context-aware implicit reminding.