2023-10-11 –, Bazar
The living heritage of traditional martial arts embodies multifaceted knowledge systems across the material and immaterial, spanning kinaesthetic, somatic, physical, social, cultural and technical ideologies of different ethnicities. Recent efforts have embarked on digitally capturing martial art performances as a foundation for knowledge preservation, yet lack efficient tools to (re)present and explore the digitised content. In particular, despite being the authentic carrier for traditional practices like martial arts, the human body has often been underrepresented and inaccessible to its knowledge.
In addressing the gap, this project inspects the combination of movement computing with ontology design to unfold kaleidoscopic knowledge dimensions in traditional martial arts. It proceeds with a dual approach: the development of a deep learning workflow to auto-classify movement series and a formal ontology conceptualising the semantic meaning of martial art movement. Integrating both allows the datafication of multimodal materials in the Hong Kong Martial Arts Living Archive (HKMALA), relating feature-based classification to semantic representation. On that basis, it instantiates an interactive knowledge system to allow users to investigate archival content with ontology-based knowledge representation and through interactive exploration via semantic and embodied clues.
This presentation will outline the methodology and showcase a series of computational experiments with the HKMALA data. Furthermore, the speaker will reflect on the notion of embodiment and how the conceptualisation and operationalisation of it may forge a new paradigm for archival interaction, facilitating the valorisation and dissemination of the intangible cultural heritage embodied.
Yumeng Hou is a digital humanities researcher currently pursuing a PhD at the Laboratory for Experimental Museology (eM+), EPFL. Trained as a data scientist, Yumeng holds experience in data visualisation, machine learning, analytics, and cloud solutions. Her research examines the convergence of cultural heritage, computational archives, digital narratives, knowledge encoding and cultural AI. Yumeng is passionate about leveraging computational strategies to augment cultural studies, especially in the fields of intangible cultural heritage and traditional folklore. She is preparing a thesis on the datafication and (re)presentation of embodied knowledge in Southern Chinese martial arts in (digital) museum settings.