By collecting (applicable) data and metrics from the conference, an algorithm could diagnose progress toward predefined goals.
The type of data could include the social interactions around the proverbial water-cooler. It might include the number of returning visitors to online fora who make contributions to set topics.
Data about person interaction like this is discussed within Social Physics.
Evaluating the quality, quantity and applicability of such various activity would be parsed using models informed by the I-Space model.
This diagnostic indicator could (automatically) trigger interventions based on Good Practice or otherwise, as indicated in the Cynefin framework.