FastPSL: A Fast Probabilistic Soft Logic Method for Logic Reasoning with Long Structured Rules in Large Knowledge Bases

发布者:刘茜茜发布时间:2026-04-13浏览次数:10

江苏省应用数学(中国矿业大学)中心系列学术报告


报告题目:FastPSL: A Fast Probabilistic Soft Logic Method for Logic Reasoning with Long Structured Rules in Large Knowledge Bases

人:蔡云峰 教授 单位: 北京雁栖湖应用数学研究院

报告时间:2026414日(周二)晚上 20:00-21:00

腾讯会议:188-920-818

欢迎全校师生参加!

                          数学学院

告人及报告内容简介:

蔡云峰于2000年至2004年就读于中国科学技术大学数学专业,20091月在北京大学获得计算数学博士学位。20091月至20126月,他先后在中国科学院数学与系统科学研究院和加州大学戴维斯分校从事博士后研究。20129月至20189月,他在北京大学担任研究员。20189月加入百度研究院,任研究科学家,至20245月离职。20246月起,他就任北京雁栖湖应用数学研究院(BIMSA)教授至今。他的研究兴趣包括机器学习、人工智能驱动的科学(AI for Science),以及人工智能与计算数学的交叉领域。

Abstract: Probabilistic soft logic (PSL) is a framework for modeling probabilistic and relational domains, and its achieves state-of-the-art results in many areas, e.g. social-network analysis, recommender system, etc. However, PSL does not scale for large databases due to its O(\sum_j \ell_j|\mathcal{E}|^{n_j}) computational complexity, where \ell_j and n_j are the length and the number of entity variables of the $j$-th rule, respectively, |\mathcal{E}| is the number of entities.In this work, we propose FastPSL, a fast logic reasoning method under the PSL framework, designed for large knowledge bases with long structured rules.FastPSL reduces the complexity to O(\sum_j \ell_jn_j|\mathcal{E}|^3). Numerical experiments on both synthetic and real-world knowledge bases demonstrate that FastPSL achieves competitive results compared to other rule-based reasoning method, while showing better performance than other knowledge graph completion baseline over long and complex rule reasoning.