Many art forms and media benefit from streaming, as it allows the consumer experience to change with artist updates and places constraints on pirates. However, these piracy constraints are superficial if the pirate can obtain a perfect copy of the source content by simply recording a stream.
The danger of piracy can be greatly mitigated if the source content is designed in such a way that each stream is slightly different. Ideally, the difficulty of reconstructing the source content would increase exponentially with the artist's work and the source content's complexity.
A model with probabilistic streaming lends itself to interactivity, as a consumer's preferences can guide the construction of each stream, and this introduces additional complexity for adversaries to reconstruct. Interactivity effectively multiplies the pirate's required sample size by the complexity of the consumer interaction, so that the reconstruction can appeal to a variety of consumer preferences.
This repository contains both experimental models for interactive content networks and analysis using tools from the field of data privacy.
- Created repository
- Initialized with model and paper from Caltech Data Privacy project