CAnDE: Entity‐Based Relationship Exploration Using a Cards and Decks Metaphor

Abstract Traditionally, police investigators structure and analyze evidence by pinning them to (cork) boards, called evidence boards. This enables investigators to group different pieces of information to one entity (e.g., person or company) and establish links between entities, commonly performed by spanning (red) threads between them. With the increasing amount of digital evidence, analysis shifts to visual analytics approaches. Current visual analytics approaches assume the evidence network is known beforehand and focus on the visualization. Furthermore, the evidence network is typically assumed static and there is limited support for rich data types. This lack of flexibility prevents investigators from identifying, forming and discovering critical entities and their relationships, potentially leading to unsolved crimes. We present CAnDE, a visual analytics approach inspired by evidence boards to support investigators in forming and exploring entities and their relationships relationships in digital criminal data. CAnDE represents dynamic entities as cards and supports interactions similar to physical actions, such as flipping, rotating and stacking. Furthermore, CAnDE extends real‐world evidence boards by combining stacked cards to a deck that allows for (1) an aggregated view of the information of all underlying entities and (2) flexible arrangement. We demonstrate the effectiveness of CAnDE by applying it to two intelligence scenarios and conducting interviews with 26 national police investigators.