Figure 4

Weights automatically determined for each dataset contributing to predictions in each context. Weights are calculated as the influence of each dataset on the posterior probability in the process or development network's Bayesian classifier, where a higher number indicates a greater influence.