Visual Representation of Natural Language Scene Descriptions

We are mainly interested in the development of CAD systems for architectural design. In the last few years CAD systems have been evolving from simple drafting tools to much more complex solid modeling environments. Nevertheless, experience has shown that an effective use of such systems relies to a large extent on the characteristics of their user interface: the interaction language should be easy to learn and use while, at the same time, it should leave the possibility of using the full power of the system. In particular, the user should have the possibility of describing in full detail a particular scenario or giving the system only a raw description of it. This paper describes NALIG, a system able to ‘understand’ and ‘reason about’ high level descriptions of spatial scenes. The user interacts with the system by using a simple natural language fragment but expressive enough to describe complex configurations of objects. In response to the user’s input NALIG builds and updates a ‘mental image’ of the scenario and replies by drawing on the screen an image mirroring its own ‘understanding’ of the scene described. The comprehension process involves various forms of common sense reasoning carried out at different levels of abstraction. This has required the integration of different AI-techniques (e.g. natural language understanding, spatial reasoning, default reasoning). We believe that NALIG is to be regarded as a significant attempt toward the development of intelligent systems capable to ‘reason about’ user specifications rather than ‘merely executing’ imperative commands