基于语义和结构置信度的知识图谱质量校验方法

doi:10.19678/j.issn.1000-3428.0064148

摘要/Abstract

摘要： 知识图谱因其较强的表达能力和可解释性而被广泛应用于问答系统、信息检索等人工智能任务中，然而，在实际应用场景中大量使用自动化知识图谱构建技术会不可避免地引入噪声和冲突，从而对知识图谱下游应用的性能产生严重影响。为从知识图谱中检测出潜在的噪声、保存真实可信的三元组并为下游应用任务提供高质量的知识，提出一种基于语义与结构双重置信度的三元组评估模型。该模型由语义真实性评估器与结构真实性评估器构成，前者通过特定规则将三元组转换为句子序列，基于双向编码器表示变换模型度量语义真实性，后者通过表示学习模型获取实体及关系的向量表示，在知识表示、路径特征两个层面上度量结构真实性。在4个真实图谱数据集上的实验结果表明，所提模型的准确率、精确率、召回率、F1值等评估指标相较TransE-RFC、TransE-KNC、TransE-XGB等模型提升3%~4%，其能够有效检测带噪声图谱数据集中的噪声错误同时最大程度地保留真实可信的知识。

关键词: 知识图谱, 质量校验, 三元组置信度评估, 语义真实性, 结构置信度

Abstract: Knowledge graphs are widely used in artificial intelligence tasks such as question-answering systems and information retrieval owing to their strong expressive ability and interpretability.The extensive utilization of automated knowledge graph construction technology in practical scenarios introduces noise and conflict and has a serious impact on the performance of downstream applications of knowledge graphs.A triplet evaluation model is proposed based on semantic and structural double trustworthiness to detect potential noise from knowledge graph，preserve the authentic and credible triples，and provide high-quality knowledge for downstream application tasks.The model comprises semantic and structural authenticity evaluators.The former converts triples into sentence sequences through specific rules and measures semantic authenticity based on the Bidirectional Encoder Representations from Transformers（BERT） model.The latter measures structural authenticity at two levels，namely，knowledge representation and path feature，by vector representation of entities and relationships obtained from the learning model. Results from analysis of four real graph datasets indicate that the accuracy，precision，recall，F1 value，and other evaluation indicators of the proposed model are 3% to 4% higher than those of TransE-Random Forest Classifier（TransE-RFC），TransE-K Nearest Neighbor Classifier （TransE-KNC），TransE-eXtreme Gradient Boosting（TransE-XGB），and other models. The proposed model can effectively detect noise errors in graph datasets while preserving authentic and credible knowledge to the maximum extent.

Key words: knowledge graph, quality verification, triple trustworthiness evaluation, semantic authenticity, structural trustworthiness

中图分类号:

TP18

叶琪, 张一乾, 阮彤, 杜渂. 基于语义和结构置信度的知识图谱质量校验方法[J]. 计算机工程, 2023, 49(5): 48-55.

YE Qi, ZHANG Yiqian, RUAN Tong, DU Wen. Quality Verification Method for Knowledge Graph Based on Semantic and Structural Trustworthiness[J]. Computer Engineering, 2023, 49(5): 48-55.

https://www.ecice06.com/CN/Y2023/V49/I5/48

图/表 10

20230515184820

20230515184824

20230515184827

20230515184833

20230515184836

20230515184839

20230515184843

20230515184846

20230515184850

20230515184854

参考文献

[1] ZOU X H.A survey on application of knowledge graph[J].Journal of Physics:Conference Series,2020,1487(1):012016.
[2] 王鑫,傅强,王林,等.知识图谱可视化查询技术综述[J].计算机工程,2020,46(6):1-11. WANG X,FU Q,WANG L,et al.Survey on visualization query technology of knowledge graph[J].Computer Engineering,2020,46(6):1-11.(in Chinese)
[3] ZHANG Y Y,DAI H J,KOZAREVA Z,et al.Variational reasoning for question answering with knowledge graph[J].Proceedings of the AAAI Conference on Artificial Intelligence,2018,32(1):6069-6076.
[4] ZHOU K,ZHAO W X,BIAN S Q,et al.Improving conversational recommender systems via knowledge graph based semantic fusion[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.New York,USA:ACM Press,2020:1006-1014.
[5] BOUNHAS I,SOUDANI N,SLIMANI Y.Building a morpho-semantic knowledge graph for Arabic information retrieval[J].Information Processing & Management,2020,57(6):102124.
[6] PAULHEIM H.Knowledge graph refinement:a survey of approaches and evaluation methods[J].Semantic Web,2016,8(3):489-508.
[7] JAYAWARDENE V,SADIQ S,INDULSKA M.An analysis of data quality dimensions[EB/OL].[2022-02-05].https://core.ac.uk/download/pdf/18442961.pdf.
[8] WANG X Y,CHEN L Z,BAN T Y,et al.Knowledge graph quality control:a survey[J].Fundamental Research,2021,1(5):607-626.
[9] SPERANSKAYA M,SCHMITT M,ROTH B.Ranking vs.classifying:measuring knowledge base completion quality[EB/OL].[2022-02-05].https://arxiv.org/abs/2102.06145.
[10] 张晓明,孙维雅,王会勇.基于知识表示学习的知识可信度评估[J].计算机工程,2021,47(7):44-54. ZHANG X M,SUN W Y,WANG H Y.Evaluation of knowledge credibility based on knowledge representation learning[J].Computer Engineering,2021,47(7):44-54.(in Chinese)
[11] LI Y F,LI X Y,LEI M J.CTransE:an effective information credibility evaluation method based on classified translating embedding in knowledge graphs[EB/OL].[2022-02-05].https://link.springer.com/chapter/10.1007/978-3-030-59051-2_19.
[12] HONG Y,BU C Y,WU X D.High-quality noise detection for knowledge graph embedding with rule-based triple confidence[EB/OL].[2022-02-05].https://link.springer.com/chapter/10.1007/978-3-030-89188-6_43.
[13] BOUGIATIOTIS K,FASOULIS R,AISOPOS F,et al.Guiding graph embeddings using path-ranking methods for error detection innoisy knowledge graphs[EB/OL].[2022-02-05].https://arxiv.org/abs/2002.08762.
[14] XIE R B,LIU Z Y,LIN F,et al.Does William Shakespeare REALLY write Hamlet? knowledge representation learning with confidence[J].Proceedings of the AAAI Conference on Artificial Intelligence,2018,32(1):1-8.
[15] JIA S B,XIANG Y,CHEN X J,et al.Triple trustworthiness measurement for knowledge graph[C]//Proceedings of WWW'19.New York,USA:ACM Press,2019:2865-2871.
[16] CUI Y,CHE W,LIU T,et al.Pre-training with whole word masking for Chinese BERT[EB/OL].[2022-02-05].https://arxiv.org/abs/1906.08101.
[17] DEVLIN J,CHANG M,LEE K,et al.BERT:pre-training of deep bidirectional transformers for language understanding[EB/OL].[2022-02-05].https://arxiv.org/abs/1810.04805.
[18] BORDES A,USUNIER N,GARCIA-DURAN A,et al.Translating embeddings for modeling multi-relational data[EB/OL].[2022-02-05].https://papers.nips.cc/paper/2013/file/1cecc7a77928ca8133fa24680a88d2f9-Paper.pdf.
[19] CHANG D J,CHEN M S,LIU C Z,et al.DiaKG:an annotated diabetes dataset for medical knowledge graph construction[EB/OL].[2022-02-05].https://arxiv.org/pdf/2105.15033.pdf.
[20] WU Y T,LIU X,FENG Y S,et al.Relation-aware entity alignment for heterogeneous knowledge graphs[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence.Washington D.C.,USA:IEEE Press,2019:5278-5284.
[21] PAL M.Random forest classifier for remote sensing classification[J].International Journal of Remote Sensing,2005,26(1):217-222.
[22] LIAO Y H,VEMURI V R.Use of K-Nearest Neighbor classifier for intrusion detection[J].Computers & Security,2002,21(5):439-448.
[23] CHEN T,HE T,BENESTY M,et al.xgboost:eXtreme gradient boosting[EB/OL].[2022-02-05].http://ftp.heanet.ie/mirrors/cran.r-project.org/web/packages/xgboost/vignettes/xgboost.pdf.
[24] DAVIS J,GOADRICH M.The relationship between Precision-Recall and ROC curves[C]//Proceedings of the 23rd International Conference on Machine Learning.New York,USA:ACM Press,2006:233-240.
[25] OZENNE B,SUBTIL F,MAUCORT-BOULCH D.The Precision-Recall curve overcame the optimism of the receiver operating characteristic curve in rare diseases[J].Journal of Clinical Epidemiology,2015,68(8):855-859.

选择文件类型/文献管理软件名称

选择包含的内容