The case study of the system by Jordanous maps the computational serendipity framework to the example interactive genetic music system somewhat loosely. The "bridge", described earlier as "reasoning techniques and/or social or otherwise externally enacted alternatives to create a bridge from the trigger to a result", is mapped to the genetic operators implemented in the system, which do not seem A) reasoning-based, B) social or C) external — they are an update in an iterative search process. Similarly, "curiosity" is mapped to the drive to model users' preferences through their ratings of novelty and appropriateness, but it is difficult to argue that the system is being curious here — it is just trying to follow what the human (who may be acting curiously) is seeking. The flexibility with which these terms have been applied to existing CC systems does not help show the uniqueness of the model — couldn't most CC systems be mapped similarly? This reviewer would prefer to
see the authors critique what the two systems they use as case studies do not do, by drawing the equivalencies they currently draw and then highlighting what is missing. This would highlight how the serendipity model can be used to show possible improvements to existing systems.
All of this is done, and I've tweaked the definition and SPECS criterion slightly to make it clear that the bridge isn't limited to (A-C) above but can use any computational method.
The case study of the system by Jordanous maps the computational serendipity framework to the example interactive genetic music system somewhat loosely. The "bridge", described earlier as "reasoning techniques and/or social or otherwise externally enacted alternatives to create a bridge from the trigger to a result", is mapped to the genetic operators implemented in the system, which do not seem A) reasoning-based, B) social or C) external — they are an update in an iterative search process. Similarly, "curiosity" is mapped to the drive to model users' preferences through their ratings of novelty and appropriateness, but it is difficult to argue that the system is being curious here — it is just trying to follow what the human (who may be acting curiously) is seeking. The flexibility with which these terms have been applied to existing CC systems does not help show the uniqueness of the model — couldn't most CC systems be mapped similarly? This reviewer would prefer to see the authors critique what the two systems they use as case studies do not do, by drawing the equivalencies they currently draw and then highlighting what is missing. This would highlight how the serendipity model can be used to show possible improvements to existing systems.