My PhD was about interactive and non-interactive evolutionary control of sound synthesis.
I've just started a 2-year IRCSET-funded postdoc in the NCRA group in UCD. I'm studying evolution and grammars, applied to architecture/shape design and musical composition.
There are three strands to this research. The first concerns the grammatical representation itself. New grammars, grammatical encodings, and techniques for improving evolutionary performance are being investigated. They are tested on a number of benchmark problems. A central aim is to produce a continuous encoding which leads to well-behaved operators, including crossover and interpolation.
A second strand of research is about evolutionary design. The aim is to create a new tool for architectural design which uses interactive evolution to generate and refine 3d objects (even buildings), rather than requiring the user to create each building manually. The representation uses shape grammars, a type of formal grammar allowing the specification of families of abstract shapes. Interactive grammatical evolution allows the user to explore the space of possible shapes defined by a grammar customised to the project. Preliminary work will be presented at GECCO 09. The encoding and interpolation operator developed in strand 1, above, will be used here to make interactive evolution more effective.
The final strand is about evolutionary and generative music. Music is amenable to both analysis and generation by formal grammars, and so is a natural application domain for grammatical evolution research. Again, the grammatical representations developed in strand 1 will be applied here. One (early) component of this is the Elevated Pitch project, in which listeners compare melodies generated by grammatical evolution so that we can study the fitness functions and grammars which lead to good music. Another component is the Mutant Turtle Music project, where grammatical formalism usually applied to generative graphics is applied to music instead, with some fun results. This is a work in progress.
I'm also collaborating on research in applying grammatical evolution to game theory and the evolution of game-playing agents, and in computational models of language evolution.