Abstract: In this talk we investigate the inference processes for probabilistic evidence. We will look into generalisation of two well studied inference processes (Maximum Entropy and Centre of Mass) to first order languages and will investigate the complexity of evidence sets for which such generalisations are possible.