EncodingActs investigates how computational methods can facilitate the transmission of multifaceted knowledge within intangible cultural heritage (ICH). ICH, characterized by living, tacit, yet complex epistemic systems, is conventionally viewed as challenging to archive and digitally represent. This research addresses the gaps between human understanding and data science through empirical and analytical studies in the context of traditional martial arts, particularly Southern Chinese martial arts. With a focus on modeling and representing embodied aspects, it enhances the transfer of martial arts knowledge from the Hong Kong Martial Arts Living Archive to potential learners.

This dissertation introduces an integrated methodology based on two main pillars: ontological datafication and motion computation. At its core is the formalization of a Martial Arts Ontology, which transforms multimodal documentation into structured data and organizes martial concepts into knowledge graphs. Machine learning techniques are utilized to translate the motion aspects of martial arts into quantifiable features and operational units. By combining these approaches, a knowledge-centric encoding strategy is developed, encompassing concepts, motion dynamics, ideologies, social agents, and interpretations, all within the context of traditional semantics. The encoding methods have enabled the development of an interactive system that makes martial arts knowledge visual, searchable, and interpretable, with its effectiveness assessed through user testing.

In summary, this dissertation examines a viable methodology and practical tools for encoding embodied knowledge in traditional Chinese martial arts. The models and tools have undergone performance evaluation and are openly available for reapplication. The findings highlight the potential of transforming ICH archives into educational materials and emphasize the importance of contextual examination in developing effective digital humanities applications.