"Big picture" predicts destination attractiveness: The role of physical breadth and contextual breadth

In global tourism, renowned attractions with diverse visual styles consistently yield positive experiences. This study introduced the 'big picture' metaphor as a universal visual code underlying their appeal. Drawing on the broaden-and-build theory, we proposed a two-dimensional visual breadth (2DVB) model, identifying physical breadth (expansiveness of visual fields) and contextual breadth (variety of visual contexts) as key predictors of destination ratings. A deep neural network with advanced feature recognition was developed to operationalize this model. Analyzing 588,821 photos from 120 global destinations, our analyses showed that both the physical and contextual visual breadth positively predicted destination ratings, validating the model. This approach surpassed traditional content-based methods, offering a new framework for cross-scene analysis in tourism management, guiding strategic planning and promotion.