Connected Spirituality
Combining electroacoustic music improvisation and abstract painting, this research explores how sound and image can express the same emotion, movement, and energy.
It introduces the ideas of Connected Nature and Connected Spirituality — showing how art can unite technology, emotion, and the natural world.
Through reflection and creative analysis, the study reveals how technology can deepen our connection to nature.
The final performance demonstrates how music and painting can merge into one expressive language, creating a more integrated and human approach to creativity.
Carefully Improvised Techno:
Visual Music AI
This research introduces Carefully Improvised Techno (CIT)—a hybrid, practice-led project combining electroacoustic improvisation, abstract painting, and real-time human–machine interaction. The term CIT reflects a dynamic balancing act: between control and spontaneity, human authorship and algorithmic influence, traditional artistry and new technology, the formal complexity of academic music traditions and the raw energy of the electronic music culture. By connecting sound and image spontaneously through improvisation, metaphor, and abstraction, CIT suggests new forms of visual music and performance that blend traditional artistic methods with advanced technologies. Drawing from avant-garde influences and grounded in human-centered AI and posthuman theory, the research seeks to bridge disciplines, cultures, and sensory modalities —proposing an inclusive, experimental framework for collaborative authorship in the age of intelligent machines.
A Context-Aware Semantic Music Recommendation Method
This study proposes a context-aware music recommendation framework based on four semantic features—Who, When, Where, and What (4W)—to better reflect users’ real-life listening behaviour. While existing streaming platforms prioritise collaborative filtering or artist/genre similarity, they often fail to capture the situational factors that shape personal music preferences. To address this limitation, this research analyses user-specific lifestyle patterns over a four-week period and constructs three recommendation modesSongs of the Week, Songs of the Day, and Songs of the Time—each utilising progressively more granular behavioural data.