The image of Southeast Asia : identifying a supranational image of Southeast Asia using machine learning

Abstract

The core of place marketing and branding is the concept of place image which can be defined as a group's or an individual's perception towards a particular place. Images represent a simplification of a large number of associations and pieces of information connected with the places. A level of place image that requires further investigation is supranational image - a level of a place image that is larger than a country. Therefore, the purpose of this research is to introduce a new method to identify the supranational image of Southeast Asia (SEA). This new method can help identify a unified supranational image of SEA, to transfer image from stronger countries to weaker ones, and to create regional thematic products and services. This research uses a combination of Google Cloud Vision AI and a topic modelling algorithm, latent Dirichlet allocation (LDA) to analyse 50,000 Flickr photos of the 10 countries and to identify the image of SEA which comprises beautiful and diverse landscape, rich tropical nature, elegant historic and religious architecture, bustling city scenes, and cultural and recreational activities.