Dolphin 4.0 can detect Room Types using room names, and a user can have its own set of Room Types and room names using the room dictionary. The Semantic graph detects rooms as Polyline Curves which allows it to calculate the exact corners and area of the room. Each room now can be associated with a building if every point of the room is inside a building. The current version has more Edge Types which provide users with better options to connect two rooms in Rhino3D.

Further, latest changes within the plugin [not published] can retrieve information from a Revit file through Rhino.Inside and transform it into an aGraphML, which has nested graphs. The component, which we developed in the current version, is the R2a component of the Revit_Recognizer namespace. The R2a was designed to help architects retrieve and convert the semantic data of a Revit file into an AGraphML file, which can visualise the spatial configurations in yEd. The nested graphs follow a hierarchy of site, buildings, floors and rooms. In the current version, only the door edgeType can be retrieved, and rooms cannot be associated with a building. These shortcomings are due to a lack of quality Revit files from which such information cannot be retrieved.

In future work, we envision to associate rooms with buildings and retrieve more edge Types as well as well as add levels and latest edge Types to the buildings to make sure that the architects can transform their projects completely to aGrapML files using Dolphin and Rhino3D.