Word Sense Disambiguation Based on Domain Knowledge and Word Vector Model

doi:10.13209/j.0479-8023.2017.027

Acta Scientiarum Naturalium Universitatis Pekinensis ›› 2017, Vol. 53 ›› Issue (2): 204-210.DOI: 10.13209/j.0479-8023.2017.027

• Orginal Article • Previous Articles Next Articles

Word Sense Disambiguation Based on Domain Knowledge and Word Vector Model

An YANG¹, Sujian LI^1,^2,^†(), Yun LI³

1. Key Laboratory of Computational Linguistics (Peking University), MOE, Beijing 100871
2. Collaborative Innovation Center for Language Ability, Xuzhou 221009
3. Institute of Linguistics, Chinese Academy of Social Sciences, Beijing 100732;

Received:2016-07-29 Revised:2016-09-23 Online:2017-03-20 Published:2017-03-20
Contact: Sujian LI

基于领域知识和词向量的词义消歧方法

杨安¹, 李素建^1,^2,^†(), 李芸³

1. 北京大学计算语言学教育部重点实验室, 北京 100871
2. 语言能力协同创新中心, 徐州 221009
3. 中国社会科学院语言研究所, 北京 100732

通讯作者: 李素建
基金资助:
国家自然科学基金(61273278, 61572049)资助

Abstract

Abstract:

A WSD method is presented, using domain keywords and word vector model built from unlabelled data. The effectiveness of the proposed approach is proved, compared with other WSD methods including Lesk on evaluation corpus in environmental domain. Through employing knowledge from different fields, proposed method can be adapted into the WSD task of other domains.

Key words: word sense disambiguation (WSD), word vector model, domain knowledge

摘要：

利用无标注文本构建词向量模型, 结合特定领域的关键词信息, 提出一种词义消歧方法。以环境领域的待消歧文本作为评测语料, 通过与Lesk等其他消歧方法进行比较, 证明了所提方法的有效性。通过引入不同的领域知识, 证明该方法亦可在其他领域的文本消歧任务中加以应用。

关键词: 词义消歧, 词向量模型, 领域知识

CLC Number:

TP181

An YANG, Sujian LI, Yun LI. Word Sense Disambiguation Based on Domain Knowledge and Word Vector Model[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2017, 53(2): 204-210.

杨安, 李素建, 李芸. 基于领域知识和词向量的词义消歧方法[J]. 北京大学学报自然科学版, 2017, 53(2): 204-210.

Add to citation manager EndNote|Ris|BibTeX

URL: https://xbna.pku.edu.cn/EN/10.13209/j.0479-8023.2017.027

https://xbna.pku.edu.cn/EN/Y2017/V53/I2/204

Figures/Tables 4

References 23

[1]	Agirre E, De Lacalle O L, Fellbaum C, et al. - 2010 task 17: all-words word sense disambiguation on a specific domain // Proceedings of the Workshop on Semantic Evaluations of the Association for Computational Linguistics. Singapore, 2009: 123-128
[2]	Koeling R, McCarthy D, Carroll J. Domain-specific sense distributions and predominant sense acquisition // Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Lan-guage Processing of the Association for Compu-tational Linguistics. Vancouver, 2005: 419-426
[3]	Kulkarni A, Khapra M M, Sohoney S, et al.CFILT: resource conscious approaches for all-words domain specific WSD // Proceedings of the 5th International Workshop on Semantic Evaluation of the Association for Computational Linguistics. Uppsala, 2010: 421-426
[4]	Lesk M.Automatic sense disambiguation using machine readable dictionaries: how to tell a pine cone from an ice cream cone // Proceedings of the 5th ACM Annual International Conference on Systems Documentation. New York, 1986: 24-26
[5]	Banerjee S, Pedersen T.An adapted Lesk algorithm for word sense disambiguation using WordNet // International Conference on Intelligent Text Proce-ssing and Computational Linguistics. Mexico, 2002: 136-145
[6]	Basile P, Caputo A, Semeraro G.An enhanced lesk word sense disambiguation algorithm through a distributional semantic model // COLING. Dublin, 2014: 1591-1600
[7]	Agirre E, De Lacalle O L, Soroa A, et al. Knowledge-based WSD and specific domains: performing better than generic supervised WSD // IJCAI. Pasadena, 2009: 1501-1506
[8]	Dwivedi S K, Rastogi P.Critical analysis of WSD algorithms // Proceedings of the ACM International Conference on Advances in Computing, Communica-tion and Control. Beirut, 2009: 62-67
[9]	Lee Y K, Ng H T, Chia T K.Supervised word sense disambiguation with support vector machines and multiple knowledge sources // Senseval-3: Third International Workshop on the Evaluation of Systems for the Semantic Analysis of Text. Barcelona, 2004: 137-140
[10]	Zhong Zhi, Ng H T.It makes sense: a wide-coverage word sense disambiguation system for free text // Proceedings of the ACL 2010 System Demonstrations. Uppsala, 2010: 78-83
[11]	Khapra M M, Shah S, Kedia P, et al.Domain-specific word sense disambiguation combining corpus based and wordnet based parameters // 5th International Conference on Global Wordnet (GWC2010). Mumbai, 2010: 1-9
[12]	Weeber M, Mork J G, Aronson A R.Developing a test collection for biomedical word sense disambiguation // Proceedings of the AMIA Symposium of the American Medical Informatics Association. Washington DC, 2001: 746-750
[13]	Magnini B, Strapparava C, Pezzulo G, et al.The role of domain information in word sense disambiguation. Natural Language Engineering, 2002, 8(4): 359-373
[14]	Mikolov T, Chen Kai, Corrado G, et al.Efficient estimation of word representations in vector space // ICLR. Scottsdale, 2013: 1-12
[15]	Yuan Dayu, Doherty R, Richardson J, et al. Word sense disambiguation with neural language models [EB/OL]. (2016-11-05) [2016-12-01].
[16]	Chen Xinxiong, Liu Zhiyuan, Sun Maosong.A unified model for word sense representation and disambiguation // EMNLP. Doha, 2014: 1025-1035
[17]	Miller G A.WordNet: a lexical database for English. Communications of the ACM, 1995, 38(11): 39-41
[18]	Pedersen T, Patwardhan S, Michelizzi J.WordNet:: similarity: measuring the relatedness of concepts // HLT-NAACL. Boston, 2004: 38-41
[19]	Varelas G, Voutsakis E, Raftopoulou P, et al.Semantic similarity methods in WordNet and their application to information retrieval on the web // Proceedings of the 7th annual ACM international workshop on Web information and data management. Bremen, 2005: 10-16
[20]	Bird S, Klein E, Loper E.Natural language processing with Python. Sebastopol: O’Reilly Media, 2009
[21]	Balamurali A R, Joshi A, Bhattacharyya P.Robust sense-based sentiment classification // Proceedings of the 2nd Workshop on Computational Approaches to Subjectivity and Sentiment Analysis of the Associa-tion for Computational Linguistics. Portland, 2011: 132-138
[22]	Rose T, Stevenson M, Whitehead M.The Reuters corpus Volume 1-from yesterday’s news to tomorrow’s language resources // LREC. Las Palmas, 2002: 827-832
[23]	McCarthy D, Koeling R, Weeds J, et al. Unsupervised acquisition of predominant word senses. Computa-tional Linguistics, 2007, 33(4): 553-590

消歧方法	召回率	消歧方法	召回率
random	0.205	ppr^[7]	0.285
most-frequent	0.459	IIITH1^[23]	0.528
lesk^[4]	0.203	scorer	0.546
most-similar	0.282

消歧方法	召回率	消歧方法	召回率
random	0.205	ppr^[7]	0.285
most-frequent	0.459	IIITH1^[23]	0.528
lesk^[4]	0.203	scorer	0.546
most-similar	0.282

消歧方法	召回率	消歧方法	召回率
whole	0.546	no-keyword	0.481
half-keyword	0.493	no-embedding	0.486

消歧方法	召回率	消歧方法	召回率
whole	0.546	no-keyword	0.481
half-keyword	0.493	no-embedding	0.486

词性	准确率	召回率
名词	0.579	0.574
动词	0.460	0.467

Word Sense Disambiguation Based on Domain Knowledge and Word Vector Model

基于领域知识和词向量的词义消歧方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 4

References 23

Related Articles 2

Recommended Articles

Metrics

消歧方法	召回率	消歧方法	召回率
random	0.189	most-similar	0.274
most-frequent	0.344	ppr	0.286
lesk	0.237	scorer	0.377

[1]	QIU Yu, CHENG Li, Daniyal Alghazzawi. Semantic Search on Non-Factoid Questions for Domain-Specific Question Answering Systems [J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2019, 55(1): 55-64.
[2]	XIE Zheng,WANG Teng,YONG Shanshan,CHEN Xu,SU Jiting,WANG Xin’an. A RISC CPU Oriented Reusable Functional Verification Platform Based on UVM [J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2014, 50(2): 221-227.