Bridging the Gap Between AI and Reality • Rhodes, Greece
Time: Monday, 3.11
Room: Room C
Authors: Daniel Sami Mitwalli
Abstract: In graphical languages, models often correspond to external artefacts pro- cessed by other services. To link both, language engineers usually implement import/export interfaces, while modeling itself relies on the IME’s internal rep- resentation. Translating between them requires manual triggering, creating in- consistencies, and encouraging generators to depend on internal representations. This tight coupling leads to logging into the IME. This paper introduces an ap- proach for native support of arbitrary codecs in Cinco Cloud through a custom codec interface. Implemented by language engineers, the interface defines a de- coder and encoder that map the external representation into the IME’s internal and back. This ensures both representations remain synchronized and usable after any model modification. Additionally, existing file types and semantics can be seamlessly integrated, associated files visualized, graphically and domain- specifically edited, and logging into an IME can be mitigated. Case studies in- cluding the DaoML language and the Wave platform demonstrate the practical applicability of this approach, showing how existing languages can be efficiently integrated to advance language engineering.