Agents in the era of big data: what the "end of theory" might mean for agent systems
Our ability to collect, manage and analyze vast amounts of data has led some to predict the demise of theory. This has important implications for research in agent systems. It can mean that specifications of agent intent, or of agent behaviour, or the norms that constrain agent behaviour can be learnt from data and maintained in the face of continuous data streams. I will offer some examples of how the agents community is beginning to leverage data in this fashion, and what the challenges might be in the future.