地理实体与重叠空间关系联合抽取的改进CasRel模型法

doi:10.11947/j.AGCS.2023.20210722

测绘学报 ›› 2023, Vol. 52 ›› Issue (8): 1387-1397.doi: 10.11947/j.AGCS.2023.20210722

地理实体与重叠空间关系联合抽取的改进CasRel模型法

蒋萌^1,2, 杨春成^1,3,4, 尚海滨¹, 秦志龙¹, 王泽凡³

1. 中国地质大学(武汉)国家地理信息系统工程技术研究中心, 湖北武汉 430074;
2. 山东浪潮新基建科技有限公司, 山东济南 250101;
3. 中国地质大学(武汉)地质探测与评估教育部重点实验室, 湖北武汉 430074;
4. 中国地质大学(武汉)计算机学院, 湖北武汉 430074

收稿日期:2021-12-30 修回日期:2023-02-27 发布日期:2023-09-07
通讯作者: 杨春成 E-mail:yangcc@cug.edu.cn
作者简介:蒋萌(1996-),男,硕士生,研究方向为地图综合与空间数据挖掘。E-mail:jiangmgiser@cug.edu.cn
基金资助:
国家自然科学基金(42171438);地质探测与评估教育部重点实验室主任基金(GLAB2022ZR01);中央高校基本科研业务费

Improved CasRel model for joint extraction of geographic entity and overlapping space relation

JIANG Meng^1,2, YANG Chuncheng^1,3,4, SHANG Haibin¹, QIN Zhilong¹, WANG Zefan³

1. National Engineering Research for Geographic Information System, China University of Geosciences, Wuhan 430074, China;
2. Shandong Inspur New Infrastructure Technology Co., Ltd., Jinan 250101, China;
3. Key Laboratory of Geological Survey and Evaluation of Ministry of Education, China University of Geosciences, Wuhan 430074, China;
4. School of Computer Science, China University of Geosciences, Wuhan 430074, China

Received:2021-12-30 Revised:2023-02-27 Published:2023-09-07
Supported by:
The National Natural Science Foundation of China (No. 42171438); The Opening Fund of Key Laboratory of Geological Survey and Evaluation of Ministry of Education (No. GLAB2022ZR01); The Fundamental Research Funds for the Central Universities

摘要/Abstract

摘要： 地理空间文本中蕴含丰富的位置信息,这些信息对地理实体的定位提供重要支撑,地理实体和空间关系抽取是位置信息获取的关键。针对地理空间关系语料库构建问题,本文从《中国大百科全书地理分册》中以包含空间关系的句子为单位,通过标注句中空间关系,完成地理空间关系语料库的构建。针对管道式(pipeline)关系抽取模型忽略了地理实体与空间关系之间的相关性问题,本文采用融合ERNIE(enhanced representation through knowledge integration)预训练模型和BAB(BiLSTM+self-attention mechanism+BiLSTM)模块改进CasRel模型实现地理实体与空间关系的联合抽取,并通过级联标注的方式解决地理空间文本中重叠空间关系的抽取。试验表明,在DuIE数据集和本文构建的地理空间关系语料库上,相较于CasRel联合抽取模型,本文模型的F₁值分别提升了4.81%和1.97%,并有效地提升重叠空间关系的抽取效果。

关键词: 空间关系抽取, ERNIE, 地理实体, 地理空间关系语料库, 重叠关系

Abstract: Geospatial text contains rich location information, which provides important support for the location of geographic entities. The extraction of geographic entities and spatial relationships is the key to obtaining location information. Aiming at the construction of the geospatial relation corpus, we take the sentences containing the spatial relation as the unit from the Encyclopedia of China Geography, and complete the construction of the geospatial relation corpus by marking the spatial relation in the sentence. For the pipeline relation extraction model which ignores the correlation between geographic entities and spatial relations, we use enhanced representation through knowledge integration (ERNIE) and BiLSTM+self-attention mechanism+BiLSTM (BAB) layers to improve the CasRel model to achieve joint extraction of geographic entities and spatial relationships, and solve the extraction of overlapping spatial relationships in geospatial texts by cascading annotation. Experiments show that on the DuIE dataset and our constructed geospatial corpus, compared with the CasRel joint extraction model, the F₁ value of our model is increased by 4.81% and 1.97%, respectively, and the extraction effect of overlapping spatial relationships is effectively improved.

Key words: spatial relation extraction, ERNIE, geographic entities, corpus of geospatial relations, relationship overlap

中图分类号:

P227

蒋萌, 杨春成, 尚海滨, 秦志龙, 王泽凡. 地理实体与重叠空间关系联合抽取的改进CasRel模型法[J]. 测绘学报, 2023, 52(8): 1387-1397.

JIANG Meng, YANG Chuncheng, SHANG Haibin, QIN Zhilong, WANG Zefan. Improved CasRel model for joint extraction of geographic entity and overlapping space relation[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(8): 1387-1397.

参考文献

[1] 张雪英, 张春菊, 朱少楠.中文文本的地理空间关系标注[J]. 测绘学报, 2012, 41(3): 468-474. ZHANG Xueying, ZHANG Chunju, ZHU Shaonan. Annotation for geographical spatial relations in Chinese text[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(3): 468-474.
[2] BEKOULIS G, DELEU J, DEMEESTER T, et al. Joint entity recognition and relation extraction as a multi-head selection problem[J]. Expert Systems with Applications, 2018, 114: 34-45.
[3] MIWA M, SASAKI Y. Modeling joint entity and relation extraction with table representation[C]//Proceedings of 2014 Conference on Empirical Methods in Natural Language Processing. Doha: Association for Computational Linguistics, 2014.
[4] KATIYAR A, CARDIE C. Going out on a limb: joint extraction of entity mentions and relations without dependency trees[C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Vancouver: Association for Computational Linguistics, 2017.
[5] MIWA M, BANSAL M. End-to-end relation extraction using LSTMs on sequences and tree structures[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Berlin: Association for Computational Linguistics, 2016.
[6] ZHENG Suncong, WANG Feng, BAO Hongyun, et al. Joint extraction of entities and relations based on a novel tagging scheme[EB/OL]. [2022-12-30]. https://arxiv.org/abs/1706.05075.
[7] BAI C, PAN L, LUO S, et al. Joint extraction of entities and relations by a novel end-to-end model with a double-pointer module[J]. Neurocomputing, 2020, 377: 325-333.
[8] ZENG Xiangrong, ZENG Daojian, HE Shizhu, et al. Extracting relational facts by an end-to-end neural model with copy mechanism[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Melbourne: Association for Computational Linguistics, 2018.
[9] BEKOULIS G, DELEU J, DEMEESTER T, et al. Joint entity recognition and relation extraction as a multi-head selection problem[J]. Expert Systems with Applications, 2018, 114: 34-45.
[10] NAYAK T, NG H T. Effective modeling of encoder-decoder architecture for joint entity and relation extraction[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(5): 8528-8535.
[11] WEI Zhepei, SU Jianlin, WANG Yue, et al. A novel cascade binary tagging framework for relational triple extraction[EB/OL]. [2022-12-30]. https://arxiv.org/abs/1909.03227.
[12] 张雪英,闾国年. 自然语言空间关系及其在GIS中的应用研究[J]. 地球信息科学, 2007, 9(6):77-81. ZHANG Xueying, LV Guonian. Natural-language spatial relations and their applications in GIS[J]. Geo-Information Science, 2007, 9(6):77-81.
[13] 袁烨城, 刘海江, 裴韬, 等. 基于语义知识的空间关系识别研究[J]. 地球信息科学学报, 2014, 16(5): 681-690. YUAN Yecheng, LIU Haijiang, PEI Tao, et al. Spatial relation extraction from Chinese characterized documents based on semantic knowledge[J]. Journal of Geo-Information Science, 2014, 16(5): 681-690.
[14] 陈功, 李霖, 邢小雨, 等. 一种规则化自然语言位置估算方法[J]. 测绘地理信息, 2015, 40(3): 39-42. CHEN Gong, LI Lin, XING Xiaoyu, et al. A method of rule-based natural language position estimation[J]. Journal of Geomatics, 2015, 40(3): 39-42.
[15] 周琦,陆叶, 李婷玉, 等. 基于语义文法的地理实体位置关系的获取[J]. 计算机科学, 2016, 43(7): 208-216. ZHOU Qi, LU Ye, LI Tingyu, et al. Acquiring relationships between geographical entities based on semantic grammar[J]. Computer Science, 2016, 43(7): 208-216.
[16] 张春菊, 张雪英, 王曙, 等. 中文文本的事件时空信息标注[J]. 中文信息学报, 2016, 30(3): 213-222. ZHANG Chunju, ZHANG Xueying, WANG Shu, et al. Annotation of spatial-temporal information of event in Chinese text[J]. Journal of Chinese Information Processing, 2016, 30(3): 213-222.
[17] ZHU Shaonan,ZHANG Xueying,ZHANG Chunju. Syntactic pattern recognition of geospatial relation description [C]//Proceedings of 2011 International Conference on Software Engineering and Multimedia Communication(SEMC 2011 V1).Qingdao:Information Engineering Research Institute,2011:366-369.
[18] 余丽,陆锋, 刘希亮. 开放式地理实体关系抽取的Bootstrapping方法[J]. 测绘学报, 2016, 45(5):616-622. YU Li, LU Feng, LIU Xiliang. A Bootstrapping based approach for open geo-entity relation extraction[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(5):616-622.
[19] 相颖, 冯钧, 夏珮珮, 等. 基于Bootstrapping的水利空间关系词提取[J]. 计算机科学, 2020, 47(12): 131-138. XIANG Ying, FENG Jun, XIA Peipei, et al. Extraction of water conservancy spatial relationship words based on bootstrapping[J]. Computer Science, 2020, 47(12): 131-138.
[20] SUN Yu, WANG Shuohuan, LI Yukun, et al. ERNIE: enhanced representation through knowledge integration[EB/OL]. [2022-11-01]. https://arxiv.org/abs/1904.09223.
[21] 刘俊楠, 刘海砚, 陈晓慧, 等. 基于地理空间数据的知识图谱构建技术研究[J]. 中文信息学报, 2020, 34(11):29-36. LIU Junnan, LIU Haiyan, CHEN Xiaohui, et al. Construction of knowledge graph based on geo-spatial data[J]. Journal of Chinese Information Processing, 2020, 34(11):29-36.
[22] 王东旭,诸云强, 潘鹏, 等. 地理数据空间本体构建及其在数据检索中的应用[J]. 地球信息科学学报, 2016, 18(4):443-452. WANG Dongxu, ZHU Yunqiang, PAN Peng, et al. Construction of geodata spatial ontology and its application in data retrieval[J]. Journal of Geo-Information Science, 2016, 18(4):443-452.
[23] 刘俊楠, 刘海砚, 陈晓慧, 等. 面向多源地理空间数据的知识图谱构建[J]. 地球信息科学学报, 2020, 22(7): 1476-1486. LIU Junnan, LIU Haiyan, CHEN Xiaohui, et al. The construction of knowledge graph towards multi-source geospatial data[J]. Journal of Geo-Information Science, 2020, 22(7): 1476-1486.
[24] 中国大百科全书总编辑委员会.中国大百科全书:中国地理[M].北京:中国大百科全书出版社, 1998. General Editorial Committee of Encyclopedia of China. Encyclopedia of China: geography of China [M]. Beijing: Encyclopedia of China Press, 1998.
[25] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all You need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach: ACM Press, 2017: 6000-6010.
[26] LI Shuangjie, HE Wei, SHI Yabing, et al. DuIE: a large-scale Chinese dataset for information extraction[M]//Natural Language Processing and Chinese Computing. Cham: Springer International Publishing, 2019: 791-800.

地理实体与重叠空间关系联合抽取的改进CasRel模型法

Improved CasRel model for joint extraction of geographic entity and overlapping space relation

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

编辑推荐

Metrics

本文评价

[1]	凌朝阳, 李锐, 吴华意, 李江, 桂志鹏. 语义驱动的地理实体关联网络构建与知识服务[J]. 测绘学报, 2023, 52(3): 478-489.
[2]	余丽, 陆锋, 刘希亮. 开放式地理实体关系抽取的Bootstrapping方法[J]. 测绘学报, 2016, 45(5): 616-622.