Integrating Logic and Probability: Algorithmic Improvements in Markov Logic Networks