‘Semantic Reality’: Situation-Aware Interactive Visualization That Shows How Objects Connect in Augmented Reality

Why Read This from an HCI Perspective

This article treats AR not as a technology that merely overlays information on a screen, but as an interaction problem: how to present relationships among multiple objects and how users can intervene. Because it visualizes relationships that commonly arise in real work—such as comparison, compatibility, and procedures—it prompts both HCI researchers and UX practitioners to think less about what ‘AI knows’ and more about what users can verify and change.

CIT's Commentary

The interesting point is that what matters is not the accuracy of the model’s guesses, but how it breaks relationships into units for display and when the user gets to check them. This is especially important in systems where safety is critical. An incorrect connection is not just a wrong answer—it also signals the user to choose the next action incorrectly. However, as the number of connecting lines grows, the screen can quickly become cluttered, so gradual disclosure that reveals only the necessary relationships first is more realistic than ‘always show everything.’ This structure is also meaningful in the context of domestic services: for mobile- and social-first products from companies like Naver and Kakao, as well as startups, the next question is how to compress and present the same principle on smaller screens than AR typically allows.

Questions to Consider While Reading

• Q.When a relationship graph becomes complex, what criteria determine which relationships to show first and which to hide?
• Q.I’m curious how quickly and clearly the system provides users with a path to modify or undo the connections it suggests.
• Q.When translating this approach to a mobile-first environment or to the context of domestic services, what interaction constraints change the most?