MediaAssist explores casual expressive browsing of personal digital media collections by reinforcing the context embedded in pictures based concepts of story-trails and dynamic tagging. Users' on-going movement through a media content space are highlighted by drawing spatial and temporal associations between content pieces through the creation of visually represented content trails. These spatial and temporal relationships are drawn based on the sequence of user interactions, which in turn rearranges the spatial content map on the display, creating visible trails joining separate pieces of media content. The trails fade over time as they are replaced by new ones, highlighting the ephemeral and changing nature of user interactions and interests. MediaAssist is being developed on the Prefuse Visualization toolkit. Force-directed graphs serve as an underlying mechanism, in which the nodes and edges are used to represent pictures and trails. These constructs serve to find and encode relationships among groups of data items, providing content and flow for picture browsing activities.