| 摘要: |
| 马尔可夫逻辑网络是将马尔可夫网络与一阶逻辑相结合的一种统计关系学习模型,在自然语言处理、复杂网络、信息抽取等领域都有重要的应用前景.较为全面、深入地总结了马尔可夫逻辑网络的理论模型、推理、权重和结构学习,最后指出了马尔可夫逻辑网络未来的主要研究方向. |
| 关键词: Markov 逻辑网 统计关系学习 概率图模型 推理 权重学习 结构学习 |
| DOI：10.3724/SP.J.1001.2011.04053 |
| 分类号: |
| 基金项目:国家自然科学基金(60970081); 国家重点基础研究发展计划(973)(2010CB327903) |
|
| Research on Markov Logic Networks |
|
XU Cong-Fu, HAO Chun-Liang, SU Bao-Jun, LOU Jun-Jie
|
|
Institute of Artificial Intelligence, Zhejiang University, Hangzhou 310027, China
|
| Abstract: |
| Markov logic networks (MLNs) is a statistical relational learning (SRL) model, which combines Markov network and first order logic. It has been applied widely in nature language processing, complex networks, information extraction, etc. This paper addresses the theoretical model of Markov logic networks (MLNs), weight and structure learning of MLNs comprehensively, and finally presents future works of MLNs. |
| Key words: Markov logic network statistical relational learning probabilistic graphical model inference weight learning structure learning |
