This presentation discusses an approach computer-aided corpus exploration that leverages machine listening and learning to discover connections between various hierarchical organisations.

As part of one of the FluCoMa plenary weeks, I presented some in progress work on software I was building for corpus exploration, using the FluCoMa tools to perform segmentation and analysis.

In this presentation I talk about the futility of modelling my own compositional desires algorithmically, and divulge an approach that uses an optimisation algorithm to express simple composition constraints with complex sonic outcomes. You can read more about the project at this link.

Kindly funded and assisted by PA Tremblay I produced a series of tutorial videos to support my work on the ReaCoMa project. For more info about the tools, see the project page to the left.