RESEARCH/ PROJECT 1.1

DYNAMIC OPTIMAL DISTINCTIVENESS IN SOUTH KOREAN FILMS

The sub-project investigates optimal distinctiveness under the contingency of time, focusing on South Korean films and using Computer Vision to elicit the novelty of films.

The sub-project explores how digitization creates forms, formats, and content. In particular, we are interested in how uniqueness of South Korean films is created, and how cultural entrepreneurs use the digital space to develop new forms and formats to create such unique products. We argue that temporal variation matters to strategically determine the novelty of films, as new films implicitly refer to prior, already existing films.

Our project is led by three research questions: How do cultural entrepreneurs (such as studios) achieve the optimal balance between similarity and distinctiveness? And how is this issue solved in industries which are essentially based on innovation and creativity, like the film industry? And finally, how do temporal dynamics affect the level of distinctiveness?

RESEARCH METHODS

This project employs a mixed-method approach. In terms of quantitative analyses, we exploit Artificial Intelligence (i.e., Computer Vision) to extract the distinctiveness of a film. We validate this measure through surveys. We then utilize the validated measure in a post-hoc econometric exercise to illustrate how similarity impacts economic and artistic success. We expect that those films are the most successful which are optimally distinct to prior and existing films. In terms of qualitative analyses, we employ semi-structured interviews to explain the underlying mechanisms at play.

The analysis on the visual dimension of Korean films provides answers to the question of how cultural entrepreneurs, especially in the context of emerging economies, utilize available resources to legitimize, on the one hand, and to be sufficiently distinctive, on the other hand.

ACTIVITIES

Natural Language Processing

NLP (Natural Language Processing, or computer-assisted text analysis) and CV (Computer Vision, or visual analysis) are two subfields within Artificial Intelligence. Both approaches provide computational methods to analyse human texts, discourses, and visual input such as images and videos. NLP...

PROJECT TEAM