This paper presents an extending inside-outside algorithm by combining the knowledge of HowNet. And then the extending algorithm is used for Chinese semantic analysis. The experiments show the extending inside-outside algorithm can overcome the shortcomings of the original one, such as high training costs, local extremum and unsatisfactory similarity to the results given by linguists.
石晶,张丽娟. 利用知网扩展内部-外部算法[J]. 现代图书情报技术, 2009, 25(7-8): 54-58.
Shi Jing,Zhang Lijuan. Extending Inside-outside Algorithm by Using HowNet. New Technology of Library and Information Service, 2009, 25(7-8): 54-58.
[1] Kakkonen T.Robustness Evaluation of Two CCG, a PCFG and a Link Grammar Parsers[C]. In:Proceedings of the 3rd Language & Technology Conference: Human Language Technologies as a Challenge for Computer Science and Linguistics, Poznan, Poland, 2007.
[2] 董振东,董强.知网[EB/OL]. [2009-03-20].http://www.keenage.com/zhiwang/c_zhiwang.html.
[3] 刘群,李素建.基于《知网》的词汇语义相似度计算[C].见:第三届汉语词汇语义学研讨会,台北.2002.
[4] Yan J. Chinese Semantic Dependency Analysis and Its Applications[D]. Tokushima:The University of Tokushima,2007.
[5] Paskin M A. Cubic-time Parsing and Learning Algorithms for Grammatical Bigram Models[R]. University of California, Berkeley, 2001.
[6] Punyakanok V, Roth D, Yih W.The Importance of Syntactic Parsing and Inference in Semantic Role Labeling[J]. Computational Linguistics, 2008,34(2):257-287.