文本推理在自然语言处理的应用中占有极为重要的位置,本文介绍了基于知网的一种推理方法,该方法以语义网络的形式表示知网中的知识,利用“标记传递”实现推理。其特点是引入构造-融合模型的思想,动态生成知识结构,有引导地在文本词汇间建立推理路径。利用16种推理类的实例对其进行测试,结果表明在有足够上下文的条件下,该方法能够得出较为理想的推理,并且代价不高。
Abstract
Text inference is central to natural language applications. This paper presents an inference method based on HowNet , which organizes knowledge with semantic net and infers with marker passing. The method introduces construction-integration model , generates knowledge structure dynamically and builds paths between text words with guide. Examples of 16 inference classes are used to test it. The results show that ideal inferences can be extracted with low cost if enough contexts are given.
关键词
计算机应用 /
中文信息处理 /
文本推理 /
构造-融合模型 /
标记传递 /
语义网
{{custom_keyword}} /
Key words
computer application /
Chinese information processing /
text inference /
construction - integration model /
marker passing /
semantic net
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] R. Schank. Language and Memory Principles. Readings in Natural Language Processing[M] . B. Grosz and K. Spark - Jones , eds. Morgan Kaufmann. 1986. 171 - 192.
[2] Charniak E. Toward a Model of Children’s Story Comprehension[R] . Artificial Intelligence Laboratory , Massachusetts Institute of Technology. AI Laboratory Technical Report 266 . 1972.
[3] Jerry R. Hobbs , Mark Stickel , Paul Martin , Douglas Edwards. Interpretation as Abduction[J] . Artificial Intelligence ,63 (1 - 2) ,1993 ,69 - 142.
[4] Hobbs , Jerry R. . Abduction in Natural Language Understanding[A] . In : L. Horn and G. Ward (Eds) , Handbook of Pragmatics , Blackwell. 2001.
[5] Schank R. C. , Abelson R. P. Scripts , plans and knowledge[A] . In :Proceedings of IJCAI 4[C] , 1975. 151 - 157.
[6] Schank R. C. , Abelson R. P. , Scripts , Plans , Goals , and Understanding[M] . Hillsdale , NJ : Erlbaum. 1977.
[7] Schank , R. C. Reminding And Memory Organization : An Introduction to MOPs[A] . In : Lehnert , W. G. ,Ringle , M. H. (Eds) Strategies for Natural Language Processing. Hillsdale : Erlbaum , 1982. 455 - 493.
[8] Norvig , P. Inference in Text Understanding[A] . In : Proceedings , AAAI - 87 ,Sixth National Conference on Artificial Intelligence[C] , Washington , July 1987. 561 - 565.
[9] Charniak , E. A Neat Theory of Marker Passing[A] . In : Proceedings , AAA - 86 , Fifth National Conference on Artificial Intelligence , Philadelphia[C] , Pennsylvania. 1986 , 584 - 588.
[10] Charniak , E. , Robert , G. A Logic for Semantic Interpretation[A] . In : Proceedings ,26th Annual Meeting of the Association for Computational Linguistics[C] . Buffalo , New York. June 1988. 87 - 94.
[11] Schank , R. C. Conceptual Information Processing[M] . North - Holland Publishing Company , 1975.
[12] Sanda M. Harabagiu , Dan I. Moldovan. A Parallel Systemfor Text Inference Using Marker Propagations[J] . IEEE Transactions on Parallel And Distributed Systems , Vol. 9 , No. 8 , August 1998.
[13] Harabagiu , Sanda M. and Dan I. Moldovan. A marker propagation text understanding and inference system[A] . In : Stewman , J. H. , ed. , Proceedings of the 9th Florida Artificial Intelligence Research Symposium[C] , Key West , 1996 , pp. 55 - 59.
[14] Walter Kintsch. The Role of Knowledge in Discourse Comprehension : A Construction-Integration Model [J] . Psychological Review.Vol. 95.No. 2. 1988. 163 - 182.
[15] 董振东,董强. 知网[H/OL] . http://www.keenage.com/zhiwang/c-zhiwang.html.
[16] Daniel Gildea-Daniel Jurafskyy. Automatic Labeling of Semantic Roles[J] . Computational Linguistics. Vol. 28. No. 3 pp. 245 - 288. September 2002.
[17] R. Hudson. Word Grammar[M] . Blackwell. Oxford. 1984.
[18] C. Rieger. Conceptual Memory and Inference. Conceptual Information Processing[A] . R. C. Schank (ed.) North - Holland , Amsterdam ,1975.
[19] 周强, 冯松岩. 构建知网关系的网状表示[J] . 中文信息学报, 2000 ,14 (6) : 21 - 27.
[20] Brill , Eric. Unsupervised Learning of Disambiguation Rules for Part-of-speech tagging[A] . Natural Language Processing Using Very Large Corpora. Kluwer Academic Publishers 1997.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金资助项目(60373056);北京市科技计划项目(H037530010830)
{{custom_fund}}
PDF(343 KB)
