1. The need for identity visualization - document graph (figure) - FenPDF (scene) 2. The means - colors are not enough (scene?) - instances of noise are bad (scene?) - modeling visual perception (figure) - design guidelines for recognizable textures (list) 3. Implementation: - design goals (list) - choosing colors (list) - basis textures (scene) - NV10 combiner (scene) - textures (scene) 3. Results and evaluation - Zipf's law (figure) - recognition experiment (list) - experiment results (table) - zoomability, scale invariance issues (show zooming) 4. Discussion (list) - no semantic correlations (out of scope for the identity visualization, could be used orthogonally to the identity visualization, if necessary) - textures of any documents similar only by chance