Associate Professor
CEMC - China Electronic Music Center
Central Conservatory of Music
Beijing China
ken@ccom.edu.cn
Abstract
Our current investigation involves three aspects of electronic music terminology: translation, conceptual or prototype modeling (ontology), and developing a useful application framework to create and maintain an online expert knowledge base. Currently, in Beijing, we are translating Curtis Roads’ Computer Music Tutorial into Chinese which in itself presents an array of challenges, both technical and logistical [1]. Using this project to define a rich though limited terminology domain, we undertake to ontologise the subject index using Protege ontology design software [2]. The working process in such a project is not at all secondary; since an ontology establishes a specialised knowledge base which is not a static entity but must be cultivated and maintained by a community, we strive to set up a public pipeline which facilitates terminology creation, annotation and syndication (via RSS). Dynamic visualisation of the evolving and self-structuring terminology topology is a future goal.
The translation of the Computer Music Tutorial into Chinese is not as straightforward as that of the French version because China has less of an academic infrastructure in this domain. There wasn’t a formalised translation for the word microsound (for example) in Chinese before we debated the issue online. The book will be used as the first comprehensive textbook in China on the subject. The issue in this case is not that of having prior audio programs without a proper textbook, but rather the fact that the textbook will in the future create new audio programs across China. In effect, it is a case of presenting (challenging) China academia with a sophisticated systematic terminology indicating the existence of a well-developed academic field which then cannot be ignored by old disciplinary regimes. The excitement of such a project derives from its aspect as, metaphorically, an undifferentiated and rapidly collapsing linguistic nebula forming objects (terms) which manifest an intellectual system and attendant research program.
The structuring of the CMT subject index into a formal ontology is, to say the least, a formidable challenge. It hasn’t been attempted before and obviously our project is an early proposal toward what must be based on a profession wide effort, rather than heroically specified by individuals. Thus, our attempt (and others as I have observed) were initially intended not to specify but to establish a community of domain experts with an interest in creating methods to specification. The first attempt at establishing this community did not produce an active project [3]. The failure may have resulted from not producing a speculative model to be debated and critiqued. It was too early in the history of semantically organised terminologies in the field of digital arts. Colleagues were initially concerned with ontology tools, software for state-tracking and discourse-mining as embedded amid wikis and listservs (Atom/RSS), or deriving methods to visualize/map such organic/emergent word nets [4].
Admittedly, our techniques are currently primitive and largely manual. First, by experiment, a core set of ontologically rich text sources are chosen. The subject index of the books are translated, put online and wikitised (made into wiki words), which yields an xml database to import into Protege. We have experimented with manual structuring of terminology classes and subclasses and the difficulties will be discussed. Using sophisticated text-analysis-to-ontology tools are a future hope which is discussed elsewhere in a comprehensive article in [5]. Our goal is in refining a rational pipeline for a professional community (such as electronic music) to foster and maintain a multilingual knowledge base. This includes experimental techniques for inside-out design in a quasi-immersive terminology environment (to be explained). Algorithms must be applied for rule-based linking of micro-ontologies into an emergent macro-ontological structure.
Though great benefits to the web masses have been predicted by semantic web visionaries, it remains to be seen if an electronic music ontology will in fact have any practical benefits for our professional community. Will music and journal databases be informed by a structured model of the knowledge domain? Will curriculum and educational programs, especially in the highly interdisciplinary field of art and technology, profit from a dynamic map of interrelated concepts? Will formalisation of concepts kill the fun for the majority of academics who fear standardisation and cling to a supposed fecund ambiguity? What makes our project unique is that a translated ontology provides possibilities for a comparative ontological analysis in the domain of sound with the possibility of discovering some mutually informing attractors.
[1] Roads, Curtis. Computer Music Tutorial. MIT Press. Cambridge. 2000.
[2] http://protege.stanford.edu
[5] Aussenac-Gilles and Sorgel (2005). Text Analysis for Ontology and Terminology Engineering. In Applied Terminology, 1(1), 35-46. IOS Press.