基于Markov逻辑网的超文本分类*

南京大学学报(自然科学版) ›› 2011, Vol. 47 ›› Issue (5): 571–577.

基于Markov逻辑网的超文本分类*

张玉芳**，孔润，田源，熊忠阳

出版日期:2015-04-27 发布日期:2015-04-27
作者简介: (重庆大学计算机学院，重庆，4ooo44)
基金资助:
重庆市自然科学基金(CSTC2008BB2191)

Hypertext classification based on Markov logic networks

Zhang Yu- Fang，Kong Run，Tian Yuan，Xiong Zhong-Yang

Online:2015-04-27 Published:2015-04-27
About author: (Department of Computer Science, Chongqing University, Chongqing, China，China)

摘要/Abstract

摘要： 在传统的监督学习任务中，实体被认为是独立同分布的.然而，现实世界中实体之间通过复杂的方式相互关联.例如在超文木分类中，具有链接关系的页面之间高度相关.标准的分类方法是忽略实
体之间的联系，对每个实体单独分类.木文将Markov逻辑网应用到超文木分类中，旨在改善这一问题. 实验结果显示了采用Markov逻辑网模型要比采用K最邻近节点算法的分类效果好;同时将实体之间
存在的联系用于学习和推理对于分类也有一定的贡献.

Abstract: In traditional supervised learning tasks，the labeled entities arc related to each other in complex ways and their labels arc not independent. For example, in hypertext classification, the labels of linked pages arc highly
correlated. A standard approach is to classify each entity independently, ignoring the corrclations between them. We use a statistically relational learning model，Markov logic networks, in hypertext classification in order to solve this
problem. Our experiments prove that this model has better performance than k-nearest neighbor does in hypertext classification and the corrclations between the entities benefit for the performance as well.

张玉芳**，孔润，田源，熊忠阳
. 基于Markov逻辑网的超文本分类*[J]. 南京大学学报(自然科学版), 2011, 47(5): 571–577.

Zhang Yu- Fang，Kong Run，Tian Yuan，Xiong Zhong-Yang
. Hypertext classification based on Markov logic networks
[J]. Journal of Nanjing University(Natural Sciences), 2011, 47(5): 571–577.

参考文献

[1]Yang Y，Slattery S, Ghani R. A study of ap proaches to hypertext categorization. Journal of intelligent Information System, 2002，18: 219一241.
[2]Craven M. Slattery S. Relational learning with statistical predicate invention; Better models for hypertext. Machine Learning, 2001，43: 97一119.
[3]Getoor L.Taskar B. introduction to statistical relational learning. Cambridge, USA:Massy chusetts Institute of Technology Press,2007 339一367.
[4]Richardson M, Dominos P. Markov logic net works. Machine Learning,2006，62:107一136.
[5]Getoor L, Fricdman N, KollerD, et al. Learn ing probabilistic relational models. Dzeroski S Lavrac N. Relational Data Mining. New York Kluwer, 2001，307一335.
[6]Wasscrman S, Faust K. Social network analy sis:Methods and applications. Cambridge UK:Cambridge University Press, 1994 342一356.
[7]Lowd D, Domingos P. Efficient weight learning for Markov logic networks. Proceedings of Principles and Practice of Knowledge Discovery in Databases, 2007，200一211.
[8]Domingos P，Richardson M. Markov logic; A unifying framework for statistical relational learning. Banff, Canada; international Confer- ence on Machine Learning，2004，49一54.
[9]Choon Y. Classification of world wide web doe umcnts. Pittsburgh:Carnegie Mcllon Universi ty, 2000.
[10]Craven M, DiPasquo D, Freitag D. Learning to extract symbolic knowledge from the world wide web. Proceedings of the Association for the Ad- vancement of Artificial Intelligence 98，1998 509一516.
[11]Taskar B, Abbecl P, Koller D. Discriminative probabilistic models for relational data. Pro- ceedings of the 18^th Conference on Uncertainty in Artificial intelligence, 2002，485一492.
[12]Kok S, Singla P，Richardson M，et al.The Al- cherry system for statistical relational Al. Technical Report, Department of Computer Sci- ence and Engineering, University of Washing- ton, Seattle, http;//alchemy, cs. Washington. edu. alchemy, cs. Washington. edu,2008.
[13]Poon H，Domingos P. Sound and efficient in- ference with probabilistic and deterministic de pendencics. Proceedings of the Association for the Advancement of Artificial intelligence，2006,458一463.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed