基于地理关联度和证据理论的地名消歧方法研究

doi:10.13209/j.0479-8023.2016.090

摘要/Abstract

摘要：

针对目前地名消歧方法普遍缺乏理论基础和统一形式化方法的现状, 以地理学第一定律为理论基础, 使用地理关联度形式化地理实体之间的邻近性。在此基础上, 提出基于证据理论的地名消歧计算模型, 用于表示与合成上下文中共现的地名证据。该模型模拟人类阅读和理解文本中时空语义的认知过程, 并为地名消歧处理提供一个统一的易扩展的形式化框架。最后, 给出本文地名消歧方法的实现算法及其实验评估。结果显示, 算法综合性能指标F1达到89.60%, 取得较好的实验效果。

关键词: 地理信息检索(GIR), 地名消歧, 地理关联度, 证据理论

Abstract: Aim

ing at the situation that previous toponym resolution researches largely lack theoretical basis and a general formal way, a concept of geo-relevance based on Tobler’s Frist Law is proposed to formalize vicinity among geographic entities. Then a toponym resolution computing model based on dempster-shafer (D-S) theory is proposed to represent and combine co-occurring toponym evidences in context. The cognitive process of human reading and understanding spatiotemporal semantics in text are simulated by D-S theory, while a general and scalable formal framework for toponym resolution is provided. Finally, an experiment evaluation is given with a good result of F1 value (89.60%).

Key words: geographic information retrieval, toponym resolution, geo-relevance, dempster-shafer theory

中图分类号:

P208
TP311

王星光, 张瑞洁, 张毅. 基于地理关联度和证据理论的地名消歧方法研究[J]. 北京大学学报自然科学版, 2017, 53(2): 344-352.

Xingguang WANG, Ruijie ZHANG, Yi ZHANG. Toponym Resolution Based on Geo-relevance and D-S Theory[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2017, 53(2): 344-352.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://xbna.pku.edu.cn/CN/10.13209/j.0479-8023.2016.090

https://xbna.pku.edu.cn/CN/Y2017/V53/I2/344

图/表 8

图1 CSGKB中的地理实体及其空间关系

Fig. 1 Geographic entities and their spatial relations in CSGKB

表1 歧义地名所有可能所指与证据源之间的定性距离(县级尺度)

Table 1 Qualitative distance among candidates belonging to ambiguous toponym and evidences (county level)

地理实体	定性地理距离
地理实体	南京市	江苏省	秦淮河	长江
南京鼓楼区	0	0	0	0
徐州鼓楼区	6	0	6	6
开封鼓楼区	12	6	12	11
福州鼓楼区	12	12	13	9

表2 歧义地名所有可能所指与证据源之间的地理关联度

Table 2 Geo-relevance among candidates belonging to ambiguous toponym and evidences

地理实体	地理关联度
地理实体	南京市	江苏省	秦淮河	长江
南京鼓楼区	1	1	1	1
徐州鼓楼区	0.0025	1	0.0025	0.0025
开封鼓楼区	6.144×10^-6	0.0025	6.144×10^-6	1.670×10^-5
福州鼓楼区	6.144×10^-6	6.144×10^-6	2.260×10^-6	0.0001

表3 证据源和证据合成的mass值

Table 3 Mass value combined by evidences

地理实体	Mass值
地理实体	南京市	江苏省	秦淮河	长江	证据合成
南京鼓楼区	0.2500	0.2500	0.2500	0.2500	0.6296
徐州鼓楼区	0.0006	0.2500	0.0006	0.0006	0.1241
开封鼓楼区	1.536×10^-6	0.0006	1.536×10^-6	4.175×10^-6	0.0003
福州鼓楼区	1.536×10^-6	1.536×10^-6	5.651×10^-7	3.085×10^-5	1.158×10^-5
未分配信度	0.7494	0.4994	0.7494	0.7493	0.2460

图2 不同证据数目下的地名消歧效果

Fig. 2 Resolution effect by different number of evidences

图3 限定窗口大小内的地名消歧效果

Fig. 3 Resolution effect within limited window of evidences

图4 组合算法的地名消歧效果

Fig. 4 Resolution effect by composition method

表4 不同地名消歧算法的性能比较

Table 4 Performance comparison of different toponym resolution methods

算法	P/%	R/%	F1/%
Baseline	70.90	70.90	70.90
RE-5	91.49	73.32	81.40
RE-19	90.42	74.89	81.92
RE_B-5	89.60	89.60	89.60

参考文献 46

[1]	Longley P A, Goodchild M F, Rhind D W.Geographic information systems and science. London: John Wiley & Sons, 2005
[2]	Smith D A, Crane G.Disambiguating geographic names in a historical digital library // Proceedings of ECDL. Darmstadt, 2001: 127-136
[3]	Amitay E, Har’El N, Sivan R, et al. Web-a-where: geotagging web content // Proceedings of the 27th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM Press, 2004: 273-280
[4]	Garbin E, Mani I.Disambiguating toponyms in news // Proceedings of the Conference on Human Language Technology and Empirical Methods in Natural Lan-guage Processing. Vancouver, BC, 2005: 363-370
[5]	Leidner J L.Toponym resolution in text: annotation, evaluation and applications of spatial grounding of place names [D]. Edinburgh: University of Edinburgh, 2007
[6]	Olligschlaeger A M, Hauptmann A G.Multimodal information systems and GIS: the informedia digital video library // Proceedings of the 1999 ESRI User Conference. San Diego, 1999: 102-106
[7]	Densham I, Reid J.A geo-coding service encom-passing a geo-parsing tool and integrated digital gazetteer service // Proceedings of the HLT-NAACL 2003 Workshop on Analysis of Geographic Referen-ces. Edmonton, 2003: 79-80
[8]	Jones C B, Purves R S.Geographical information retrieval. International Journal of Geographical Infor-mation Science, 2008, 22(3): 219-228
[9]	张毅, 王星光, 陈敏, 等. 基于语义的文本地理范围提取方法. 高技术通讯, 2012, 22(2): 165-170
[10]	Buscaldi D, Rosso P.A conceptual density-based approach for the disambiguation of toponyms. International Journal of Geographical Information Science, 2008, 22(3): 301-313
[11]	Overell S, Rüger S.Using co-occurrence models for placename disambiguation. International Journal of Geographical Information Science, 2008, 22(3): 265-287
[12]	Batista D S, Silva M J, Couto F M, et al.Geographic signatures for semantic retrieval // Proceedings of the 6th Workshop on Geographic Information Retrieval. New York: ACM, 2010: 93-100
[13]	Hu Youheng, Ge Linlin.A supervised machine learning approach to toponym disambiguation // The Geospatial Web. London: Springer, 2007: 117-128
[14]	Lieberman M D, Samet H.Adaptive context features for toponym resolution in streaming news // Pro-ceedings of the 35th International ACM SIGIR Con-ference on Research and Development in Information Retrieval. New York: ACM, 2012: 731-740
[15]	Santos J, Anastácio I, Martins B. Using machine learning methods for disambiguating place references in textual documents. Geojournal, 2015, 80(3): 375‒375
[16]	Wang Xingguang, Zhang Yi, Chen Min, et al.An evidence-based approach for toponym disambiguation // Proceedings of the Eighteenth International Con-ference on Geoinformatics. Beijing: IEEE, 2010: 1-7
[17]	Zhang Wei, Gelernter J.Geocoding location expre-ssions in Twitter messages: a preference learning method. Journal of Spatial Information Science, 2014 (9): 37-70
[18]	Li Yi, Moffat A, Stokes N, et al.Exploring proba-bilistic toponym resolution for geographical infor-mation retrieval // Proceedings of 3rd Workshop on Geographic Information Retrieval. Seattle, 2006: 17-22
[19]	Tang Xuri, Chen Xiaohe, Peng Minxuan.Toponym resolution in discourse // Natural Language Proce-ssing and Knowledge Engineering. Beijing, 2008: 1-8
[20]	Volz R, Kleb J, Mueller W.Towards ontology-based disambiguation of geographical identifiers // WWW 2007 Workshop I3. Banff, 2007: 1-7
[21]	杜萍, 刘勇. 中文地名识别与歧义消除: 以中国县级以上行政区划地名为例. 遥感技术与应用, 2011, 26(6): 868-873
[22]	唐旭日, 陈小荷, 张雪英. 中文文本的地名解析方法研究. 武汉大学学报: 信息科学版, 2010, 35(8): 930-935
[23]	Pouliquen B, Steinberger R, Ignat C, et al.Geographical information recognition and visualiza-tion in texts written in various languages // Procee-dings of the 2004 ACM Symposium on Applied Com-puting. New York: ACM, 2004: 1051-1058
[24]	Li Huifeng, Srihari R K, Niu Cheng, et al.Location normalization for information extraction // Procee-dings of the 19th International Conference on Com-putational Linguistics. Stroudsburg: Association for Computational Linguistics, 2002: 1-7
[25]	Li Huifeng, Srihari R K, Niu Cheng, et al.InfoXtract location normalization: a hybrid approach to geogra-phic references in information extraction // Procee-dings of the HLT-NAACL 2003 workshop on Analy-sis of Geographic References-Volume 1. Stroudsburg: Association for Computational Linguistics, 2003: 39-44
[26]	Overell S.Geographic information retrieval: classifi-cation, disambiguation and modelling [D]. London: Imperial College London, 2009
[27]	Hu Yingjie, Janowicz K, Prasad S. Improving Wikipedia-based place name disambiguation in short texts using structured data from DBpedia [C/OL] // Proceedings of the 8th Workshop on Geographic Information Retrieval. Dallas: ACM. (2014‒11‒04) [2015‒09‒20].
[28]	朱少楠, 张雪英, 李明, 等. 基于行政隶属关系树状图的地名消歧方法. 地理与地理信息科学, 2013, 29(3): 39-42
[29]	Bensalem I, Kholladi M K.Toponym disambiguation by arborescent relationships. Journal of Computer Science, 2010, 6(6): 653-659
[30]	Overell S, Magalhaes J, Rüger S. Place disambigua-tion with co-occurrence models // CLEF 2006 Work-shop, Working Notes. Alicante, 2006: 59-68
[31]	Overell S, Rüger S M. Identifying and grounding descriptions of places // Proceedings of the Third ACM Workshop on Geographical Information Retrie-val at SIGIR (GIR’06). Seattle, 2006: 14‒16
[32]	Leidner J L, Sinclair G, Webber B.Grounding spatial named entities for information extraction and question answering // Proceedings of the HLT-NAACL 2003 Workshop on Analysis of Geographic References-Volume 1. Stroudsburg: Association for Computational Linguistics, 2003: 31-38
[33]	Buscaldi D, Rosso P.Map-based vs. knowledge-based toponym disambiguation // Proceedings of the 2nd International Workshop on Geographic Information Retrieval (GIR’08). New York: ACM, 2008: 19-22
[34]	Lieberman M D, Samet H, Sankaranayananan J. Geotagging: Using proximity, sibling, promi-nence clues to understand comma groups // Procee-dings of the 6th Workshop on Geographic Information Retrieval (GIR’10). Zurich: ACM, 2010: 1‒8
[35]	Pouliquen B, Kimler M, Steinberger R, et al. Geo-coding multilingual texts: Recognition, disambigua-tion and visualization // Proceedings of LREC-2006. Genoa, 2006: 53‒58
[36]	Buscaldi D, Magnini B. Grounding toponyms in an Italian local news corpus // Proceedings of the 6th Workshop on Geographic Information Retrieval. Zurich, 2010: 1‒5
[37]	Buscaldi D.Approaches to disambiguating toponyms. Sigspatial Special, 2011, 3(2): 16-19
[38]	Liu Yu, Wang Fahui, Kang Chaogui, et al.Analyzing relatedness by toponym co-occurrences on web pages. Transactions in GIS, 2014, 18(1): 89-107
[39]	Tobler W R.A computer movie simulating urban growth in the Detroit region. Economic geography, 1970, 46: 234-240
[40]	Hill L.Core elements of digital gazetteers: place-names, categories, and footprints // Proceedings of the 4th European Conference on Research and Advanced Technology for Digital Libraries. Berlin: Springer, 2000: 280-290
[41]	Zhang Yi, Gao Yong, Xue Lulu, et al.A common sense geographic knowledge base for GIR. Science in China Series E: Technological Sciences, 2008, 51(1): 26-37
[42]	邬伦, 刘瑜, 张晶, 等. 地理信息系统: 原理方法和应用. 北京: 科学出版社, 2005
[43]	Longley P A, Goodchild M F, Maguire D J, et al.Geographical information systems and science. Chi-chester: John Wiley & Sons, 2005
[44]	Lloyd R, Patton D, Cammack R. Basic-level geo-graphic categories. The Professional Geographer, 1996, 48(2): 181‒181
[45]	Dempster A P.Upper and lower probabilities induced by a multivalued mapping. The Annals of Mathema-tical Statistics, 1967, 38(2): 325-339
[46]	Shafer G.A mathematical theory of evidence. Prin-ceton: Princeton University Press, 1976