WANG Jing,YU Yan,XIONG Xi.Integrating Entity Features and Latent Relation Model for Chinese Relation Extraction[J].Journal of Chengdu University of Information Technology,2025,40(03):257-263.[doi:10.16836/j.cnki.jcuit.2025.03.001]
融合实体特征和潜在关系的中文关系抽取模型
- Title:
- Integrating Entity Features and Latent Relation Model for Chinese Relation Extraction
- 文章编号:
- 2096-1618(2025)03-0257-07
- Keywords:
- relation extraction; overlap entity; potential relationships; bidirectional triplet extraction; entity feature
- 分类号:
- TP391.1
- 文献标志码:
- A
- 摘要:
- 从非结构化文本中抽取关系三元组信息对于构建知识图谱尤为重要,现有的研究方法通常以先识别实体再抽取关系为主。尽管这些方法取得了良好的性能,但忽略了实体与关系之间的内在联系,且无法有效解决同一文本中实体重叠问题。针对以上问题,提出一种融合实体特征和潜在关系的中文关系抽取模型,以关系作为条件通过主实体映射客实体。首先将实体信息以二维矩阵方式进行标记,进行主实体识别; 然后预测出文本可能存在的关系; 最后融合实体特征和潜在关系信息进行客实体识别。整个过程采用双向关系三元组抽取框架,即从两个方向上进行关系三元组的抽取,使其双向抽取结果相互补充。该模型有效保留了实体与关系之间的内在联系,增强了对重叠实体的关系识别。实验结果表明,在DuIE和CMeIE中文数据集上,提出的模型在精确率、召回率和F1评测指标上均有一定的提升,证明该模型的有效性。
- Abstract:
- Extracting relationship triad information from unstructured text is especially important for constructing knowledge graphs,and existing research methods usually focus on identifying entities before extracting relationships.Although these methods have achieved good performance,they ignore the intrinsic connection between entities and relations,and cannot effectively solve the problem of overlapping entities in the same text.To address the above problems,a Chinese relationship extraction model that integrates entity features and potential relationships is proposed,and the main idea is to map the guest entities through the main entities with the relationships as the conditions.The main idea is to map the guest entities by using the relationships as conditions.Firstly,the entity information is labeled in a two-dimensional matrix to recognize the main entities; then the possible relationships in the text are predicted; finally,the entity features,and potential relationship information are fused to recognize the guest entities.The whole process adopts a bidirectional relationship ternary extraction framework,i.e.,the relationship ternary is extracted from two directions,so that the bidirectional extraction results are complementary to each other.The model effectively preserves the intrinsic connection between entities and relationships and enhances the relationship recognition of overlapping entities.The experimental results show that the model proposed in this paper has some improvement in precision rate,recall rate,and F1 evaluation metrics on DuIE and CMeIE Chinese datasets,which proves the effectiveness of the model.
参考文献/References:
[1] 郭喜跃,何婷婷.信息抽取研究综述[J].计算机科学,2015,42(2):14-17+38.
[2] 王江鹏.基于深度学习的自然语言处理技术发展分析[J].中国安防,2022,201(12):40-43.
[3] Ratinov L,Roth D.Design challenges and misconceptions in named entity recognition[C].CoNLL ’09:Proceedings of the Thirteenth Conference on Computational Natural Language Learning.2009.
[4] Wang Z,Wen R,Chen X,et al. Finding Influential Instances for Distantly Supervised Relation Extraction[C]. International Conference on Computational Linguistics. ICCL,2022:2639-2650.
[5] Eberts M,Ulges A.Span-based Joint Entity and Relation Extraction with Transformer Pre-training[J].arXiv,2019.
[6] Zelenko D,Aone C,Richardella A.Kernel Methods for Relation Extraction[C].Empirical Methods in Natural Language Processing.Association for Computational Linguistics,2002.
[7] Chan Y S,Roth D.Exploiting Syntactico-Semantic Structures for Relation Extraction[C].The 49th Annual Meeting of the Association for Computational Linguistics:Human Language Technologies,Proceedings of the Conference,19-24 June,2011,Portland,Oregon,USA.Association for Computational Linguistics,2011.
[8] Miwa M,Sasaki Y.Modeling Joint Entity and Relation Extraction with Table Representation[C].Conference on Empirical Methods in Natural Language Processing.2014.
[9] Ren X,Wu Z,He W,et al. CoType:Joint Extraction of Typed Entities and Relations with Knowledge Bases[C].International conference on world wide web. ACM,2017:1015-1024.
[10] Zheng S,Wang F,Bao H,et al. Joint Extraction of Entities and Relations Based on a Novel Tagging Scheme[C]. Association for Computational Linguistics. ACL,2017:1227-1236.
[11] Wei Z,Su J,Wang Y,et al.A Novel Cascade Binary Tagging Framework for Relational Triple Extraction[C].Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.2020.
[12] Yu B,Zhang Z,Shu X,et al.Joint Extraction of Entities and Relations Based on a Novel Decomposition Strategy[J].arXiv,2019.
[13] Chen Y,Zhang Y,Hu C,et al.Jointly Extracting Explicit and Implicit Relational Triples with Reasoning Pattern Enhanced Binary Pointer Network[C].North American Chapter of the Association for Computational Linguistics.Association for Computational Linguistics,2021.
[14] Xuetao Tian,Liping Jing,Lu He,Feng Liu.StereoRel:Relational Triple Extraction from a Stereoscopic Perspective[C].International Joint Conference on Natural Language Processing; Annual Meeting of the Association for Computational Linguistics.2021.
[15] Devlin J,Chang M W,Lee K,et al.BERT:Pre-training of Deep Bidirectional Transformers for Language Understanding[J].arXiv,2018.
[16] 石文浩,孟军,张朋,等.融合CNN和Bi-LSTM的miRNA-lncRNA互作关系预测模型[J].计算机研究与发展,2019,56(8):1652-1660.
[17] Bohnet B,Yu J,Poesio M. Named Entity Recognition as Dependency Parsing[C].Association for Computational Linguistics. ACL,2020:6470-6476.
[18] Lin M,Chen Q,Yan S.Network In Network[Z].International Conference on Learning Representations. ICLR,2014.
[19] Lin T Y,Goyal P,Girshick R,et al.Focal Loss for Dense Object Detection[J].arXiv e-prints,2017.
[20] Li S,He W,Shi Y,et al.DuIE:A Large-Scale Chinese Dataset for Information Extraction[C].CCF International Conference on Natural Language Processing and Chinese Computing.Baidu Inc.Beijing 100193,China,2019.
[21] Guan T,Zan H,Zhou X,et al.CMeIE:Construction and Evaluation of Chinese Medical Information Extraction Dataset[C].Natural Language Processing and Chinese Computing:9th CCF International Conference.NLPCC,2020:272-282.
[22] Zeng D,Zhang H,Liu Q.CopyMTL:Copy Mechanism for Joint Extraction of Entities and Relations with Multi-Task Learning[J].Proceedings of the AAAI Conference on Artificial Intelligence,2020,34(5):9507-9514.
[23] Zeng X,He S,Zeng D,et al.Learning the Extraction Order of Multiple Relational Facts in a Sentence with Reinforcement Learning[C].Empirical Methods in Natural Language Processing.Association for Computational Linguistics,2019.
[24] Wei Z,Su J,Wang Y,et al.A Novel Cascade Binary Tagging Framework for Relational Triple Extraction[C].Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.2020.
[25] Zhang R H,Liu Q,Fan A X,et al.Minimize Exposure Bias of Seq2Seq Models in Joint Entity and Relation Extraction[C].Findings of the Association for Computational Linguistics. EMNLP,2020:236-246.
[26] 陈仁杰,郑小盈,祝永新.融合实体类别信息的实体关系联合抽取[J].计算机工程,2022,48(3):8.
备注/Memo
收稿日期:2023-10-31
基金项目:国家自然科学基金资助项目(81901389); 四川省科技计划资助项目(23ZDYF2088); 教育部人文社会科学研究基金资助项目(22YJAZH120)
通信作者:余艳.E-mail:yuyan@cuit.edu.cn