Speaker: Matthew Dimopoulos, Claremont Graduate University.  Moderator: Madeleine Clare Elish. Abstract: As an archivist and historian it is increasingly apparent that with the current and growing glut of digitally born documents a new model of visualizing collections would be helpful to researcher and archivist alike. This new visual model, when applied, should integrate and operate symbiotically with the Greene and Meissner, more product, less process model of processing archival collections. I propose an adaptation of an already familiar model in the form of a cloud of links or more precisely that of a neural network visualization complete with links akin to axons and dendrites that navigate the hierarchy of the collection. The immediate goal of this visual model is to give a dynamic three dimensional view of a singular collection along with the finding aid down to the series or folder level. However, the long term goal of this proposal is to implement the model across multiple collections thus freeing the researcher and archivist from thinking about archival collections in isolation and providing a visual method of describing the relationships across all of the various holdings of any particular institution, repository or archive. By using the existing vocabulary control inherent in EAD, this model can be adapted quickly using already existing keywords contained in the metadata of most current finding aids. This model can provide greater insight into the depth and breadth of the overall holdings as well as foster new understanding between individual collections within an archive.