Corroborating Information from Disagreeing Views

We consider a set of views stating possibly conflicting facts. Negative facts in the views may come, e.g., from functional dependencies in the underlying database schema. We want to predict the truth