This book introduces Relational Markov Networks (RMN's) for Collective Information Extraction. Also, an attempt has been made to improve the performance of Relational Markov Network using Approximate Inference Procedure such as Gibbs Sampling for Collective Information Extraction. Gibbs Sampling has been used for making the performance of LT-RMN's better than CRF's for Collective Information Extraction.