Discriminative Structure Learning of Markov Logic Networks

Marenglen Biba and Stefano Ferilli and Floriana Esposito


Markov Logic Networks (MLNs) combine Markov networks and first-order logic by attaching weights to first-order formulas and viewing these as templates for features of Markov networks. Learning the structure of MLNs is performed by state-of-the-art methods by max- imizing the likelihood of a relational database. This can lead to subop- timal results given prediction tasks. On the other hand better results in prediction problems have been achieved by discriminative learning of MLNs weights given a certain structure. In this paper we propose an algorithm for learning the structure of MLNs discriminatively by max- imimizing the conditional likelihood of the query predicates instead of the joint likelihood of all predicates. The algorithm chooses the struc- tures by maximizing conditional likelihood and sets the parameters by maximum likelihood. Experiments in two real-world domains show that the proposed algorithm improves over the state-of-the-art discriminative weight learning algorithm for MLNs in terms of conditional likelihood. We also compare the proposed algorithm with the state-of-the-art gen- erative structure learning algorithm for MLNs and con rm the results in [22] showing that for small datasets the generative algorithm is compet- itive, while for larger datasets the discriminative algorithm outperfoms the generative one.


Paper (PDF)