Learning Markov logic networks with limited number of labeled training examples

摘要: Markov Logic Networks (MLN) is a unified framework integrating first-order logic and probabilistic inference. Most existing methods of MLN learning are supervised approaches requiring large amount training examples, leading to substantial human effort for preparing these examples. To reduce such effort, we have developed semi-supervised an MLN, in particular structure from set unlabeled data limited number labeled achieve this, aim at maximizing the expected pseudo-log-likelihood function observation data, instead which commonly used MLN. evaluate our proposed method, conducted experiments on two different datasets empirical results demonstrate that effective, outperforming approach considers examples alone.