图书情报工作

图书情报工作 >

2024 , Vol. 68 >Issue 3: 105 - 116

DOI: https://doi.org/10.13266/j.issn.0252-3116.2024.03.010

知识组织

面向探索式搜索的领域知识图谱构建及实验探索

  • 王娟 ,
  • 曹树金 ,
  • 王志红 ,
  • 彭碧涛
展开
  • 1 广东外语外贸大学信息科学与技术学院 广州 510420;
    2 山东理工大学信息管理研究院 淄博 255000;
    3 中关村实验室 北京 100094
王娟,副教授,博士;王志红,助理研究员;彭碧涛,副教授,博士。

收稿日期: 2023-06-19

  修回日期: 2023-08-18

  网络出版日期: 2024-02-23

基金资助

本文系国家社会科学基金一般项目“基于深度学习的学科领域网络学术情报发现研究”（项目编号：18BTQ065）研究成果之一。

收起

Construction and Experimental Exploration of Domain Knowledge Graph for Exploratory Search

  • Wang Juan ,
  • Cao Shujin ,
  • Wang Zhihong ,
  • Peng Bitao
Expand
  • 1 School of Information Science and Technology, Guangdong University of Foreign Studies, Guangdong 510420;
    2 Institute of Information Management, Shandong University of Technology, Zibo 255000;
    3 Zhongguancun Laboratory, Beijing 102206

Received date: 2023-06-19

  Revised date: 2023-08-18

  Online published: 2024-02-23

Supported by

This work is supported by the National Social Science Fund of China titled "Research on Network Academic Information Discovery in Subject Fields Based on Deep Learning" (Grant No. 18BTQ065).

Fold

摘要

[目的/意义] 特定领域中的信息检索系统，往往因为用户自身领域知识不足存在检索效率低下的问题。现有的面向特定领域的知识组织工具，如领域知识图谱，可以有效缓解这一问题，但是如何将其更好地嵌入到现有信息检索系统中是目前尚未解决的问题。因此，提出一套基于领域知识图谱的探索式搜索系统的全流程解决方案，包括从实体及关系的联合抽取、领域知识图谱构建到探索式搜索系统的实现。[方法/过程] 以图情领域学术文献为研究对象，对图情领域的实体及其间关系的形式化定义进行深入的探索、提出基于Paddle UIE实现实体及关系的联合抽取任务方法，并构建一个基于可交互图情领域知识图谱的原型检索系统。[结果/结论] 通过比较不同实体及关系联合抽取方法的效果，包括Paddle UIE、CasRel、SpERT和CORE，发现基于提示学习的预训练大语言模型Paddle UIE具有更好的抽取效果，特别是在样本量较少的情况下。此外，从进一步设计的两个搜索实验任务结果中可以发现，与通用搜索引擎相比，本文系统可以显著提高用户满意度，有效解决用户领域知识不足（如跨学科场景）下检索性能低下的问题，表明本文提出的全流程解决方案可以为学术文献服务提供方开发用户支持工具提供流程指导和建议。

关键词： 领域知识图谱; 提示学习; 大模型; 联合抽取; 探索式搜索

本文引用格式

王娟 , 曹树金 , 王志红 , 彭碧涛 . 面向探索式搜索的领域知识图谱构建及实验探索[J]. 图书情报工作, 2024 , 68(3) : 105 -116 . DOI: 10.13266/j.issn.0252-3116.2024.03.010

Abstract

[Purpose/Significance] Information retrieval systems in specific domains often suffer from low retrieval efficiency due to the users’ lack of domain knowledge. Existing domain-specific knowledge organization tools, such as domain knowledge graphs, can effectively alleviate this problem, but how to better integrate them with existing information retrieval systems is a problem that has not been solved yet. Therefore, this paper proposes a full-process solution for implementing an exploratory search system based on domain knowledge graphs, including joint extraction of entities and their relations, domain knowledge graph construction and implementation of exploratory search system. [Method/Process] This paper focused on the academic literature in library and information science (LIS) and explored the formal definition of entities and their relations in LIS. It also proposed a joint extraction method of entities and relations on Paddle UIE, and built a prototype retrieval system on an interactive LIS knowledge graph. [Result/Conclusion] By comparing the effects of different joint extraction methods of entities and relations, including Paddle UIE, CasRel, SpERT and CORE, it finds that the pre-trained large language model Paddle UIE based on prompt learning has better extraction performance, especially in the case of small sample size. In addition, this paper further designs two search experiment tasks to analyze the effect of exploratory search system. Compared with general search engines, this system can significantly improve user satisfaction and effectively solve the problem of low retrieval performance due to insufficient user domain knowledge (such as cross-disciplinary scenarios), indicating that the full-process solution proposed in this paper can provide process guidance and suggestions for academic literature service providers to develop user support tools.

Key words： domain knowledge graph; prompt learning; large language model; joint extraction; exploratory search

参考文献

[1] MARCHIONINI G. Exploratory search:from finding to understanding[J]. Communications of the ACM, 2006, 49(4):41-46.
[2] BRIN S. Extracting patterns and relations from the World Wide Web[C]//Proceedings of the 6th international workshop on the World Wide Web and databases. Valencia:Springer, 1998:172-183.
[3] XU F, USZKOREIT H, LI H, et al. Adaptation of relation extraction rules to new domains[C]//Proceedings of the poster session of the 6th international conference on language resources and evaluation. Marrakech:European Language Resources Association, 2008:2446-2450.
[4] 杨志豪, 洪莉, 林鸿飞, 等. 基于支持向量机的生物医学文献蛋白质关系抽取[J]. 智能系统学报, 2008, 3(4):361-369. (YANG Z H, HONG L, LIN H F, et al. Extraction of information on protein-protein interaction from biomedical literatures using an SVM[J]. CAAI transactions on intelligent systems, 2008, 3(4):361-369.)
[5] 何婷婷, 徐超, 李晶, 等. 基于种子自扩展的命名实体关系抽取方法[J]. 计算机工程, 2006, 32(21):183-184. (HE T T, XU C, LI J, et al. Named entity relation extraction method based on seed self-expansion[J]. Computer engineering, 2006, 32(21):183-184.)
[6] ZHANG D, WANG D. Relation classification via recurrent neural network[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1508.01006.pdf.
[7] ZENG D, LIU K, LAI S, et al. Relation classification via convolutional deep neural network[C]//Proceedings of the 25th international conference on computational linguistics. Dublin:Dublin City University and Association for Computational Linguistics, 2014:2335-2344.
[8] XU Y, MOU L, LI G, et al. Classifying relations via long short term memory networks along shortest dependency path[J]. Computer science, 2015, 42:56-61.
[9] EBRAHIMI J, DOU D. Chain based RNN for relation classification[C]//Proceedings of the 2015 conference of the North American chapter of the Association for Computational Linguistics. Denver:Association for Computational Linguistics, 2015:1244-1249.
[10] LI Z, YANG Z, SHEN C, et al. Integrating shortest dependency path and sentence sequence into a deep learning framework for relation extraction in clinical text[J]. BMC medical informatics and decision making, 2019, 19(1):43-50.
[11] 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. Berlin:Association for Computational Linguistics, 2016:1105-1116.
[12] ZHENG S, WANG F, BAO H, et al. Joint extraction of entities and relations based on a novel tagging scheme[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1706.05075.pdf.
[13] WEI Z, SU J, WANG Y, et al. A novel cascade binary tagging framework for relational triple extraction[EB/OL].[2023-11-01]. https://aclanthology.org/2020.acl-main.136.pdf.
[14] EBERTS M, ULGES A. Span-based joint entity and relation extraction with transformer pre-training[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1909.07755.pdf.
[15] LIU P, YUAN W, FU J, et al. Pre-train, prompt, and predict:a systematic survey of prompting methods in natural language processing[EB/OL].[2023-11-01]. https://arxiv.org/pdf/2107. 13586.pdf.
[16] XU Y, JONES G, WANG B. Query dependent pseudo relevance feedback based on Wikipedia[C]//Proceedings of the 32nd international ACM SIGIR conference on research and development in information retrieval. Boston:Association for Computing Machinery, 2009:59-66.
[17] DALTON J, DIETZ L, ALLAN J. Entity query feature expansion using knowledge base links[C]//Proceedings of the 37th international ACM SIGIR conference on research & development in information retrieval. Gold Coast:Association for Computing Machinery, 2014:365-374.
[18] FUKAZAWA Y, KARAPETSAS E, ZHU D, et al. Exploratory activity search[J]. International journal of knowledge-based and intelligent systems, 2015, 19(1):15-25.
[19] 杨倩. 探索式搜索行为的先验知识分析与信息服务策略研究[J]. 图书情报知识, 2021(2):144-153. (YANG Q. Research on priori knowledge analysis and information service strategy of exploratory search behavior[J]. Documentation, information & knowledge, 2021(2):144-153.)
[20] 邢玉艳, 刘萍. 探索式搜索前后用户认知结构变化研究[J]. 图书情报工作, 2021, 65(22):74-84. (XING Y, LIU P. Research on users' cognitive structure changes before and after exploratory search[J]. Library and information service, 2021, 65(22):74-84.)
[21] LU Y, LIU Q, DAI D, et al. Unified structure generation for universal information extraction[EB/OL].[2023-11-01]. https://arxiv.org/pdf/2203.12277.pdf.
[22] SUN Y, WANG S, FENG S, et al. ERNIE 3.0:large-scale knowledge enhanced pre-training for language understanding and generation[EB/OL].[2023-11-01]. https://arxiv.org/pdf/2107.02137.pdf.
[23] 黄恒琪, 于娟, 廖晓, 等. 知识图谱研究综述[J]. 计算机应用, 2019, 28(6):1-12. (HUANG H Q, YU J, LIAO X, et al. Review on knowledge graphs[J]. Computer systems & applications, 2019, 28(6):1-12.)
[24] AUER S, BIZER C, LEHMANN J, et al. DBpedia:a nucleus for a web of open data[C]//Proceedings of the 6th international semantic web conference and 2nd Asian semantic web conference. Busan:Springer, 2007:715-728.
[25] NIU X, SUN X, WANG H, et al. Zhishi.me-weaving Chinese linking open data[C]//Proceedings of the 10th international semantic web. Berlin:Springer, 2011:205-220.
[26] XU B, XU Y, LIANG J, et al. CN-DBpedia:a never-ending Chinese knowledge extraction system[C]//Proceedings of the 30th international conference on industrial engineering and other applications of applied intelligent systems. Arras:Springer, 2017:428-438.
[27] ROTH D, YIH W. A linear programming formulation for global inference in natural language tasks[C]//Proceedings of the 8th conference on computational natural language learning at HLTNAACL. Boston:Association for Computational Linguistics, 2004:1-8.
[28] LUAN Y, HE L, OSTENDORF M, et al. Multi-task identification of entities, relations, and conference for scientific knowledge graph construction[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1808.09602.pdf.
[29] AUGENSTEIN I, DAS M, RIEDEL S, et al. SemEval 2017 task 10:ScienceIE-extracting keyphrases and relations from scientific publications[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1704.02853.pdf.
[30] GURULINGAPPA H, RAJPUT A, ROBERTS A, et al. Development of a benchmark corpus to support the automatic extraction of drug-related adverse effects from medical case reports[J]. Biomedical informatics, 2012, 45(5):885-892.
[31] 张瑞. 我国图书情报学跨学科知识流入特征研究[J]. 情报杂志, 2019, 38(8):195-201. (ZHANG R. Study on the interdisciplinary knowledge inflow characteristics of library and information science in China[J]. Journal of intelligence, 2019, 38(28):195-201.)
[32] WANG Y, HU J. A machine learning based approach for table detection on the web[C]//Proceedings of the 11th international conference on World Wide Web. Honolulu:Association for Computing Machinery, 2002:242-250.
[33] RAMESH B, SETHI R, YU H, et al. Figure-associated text summarization and evaluation[EB/OL].[2023-11-01]. https://doi.org/10.1371/journal.pone.0115671.
[34] SIEGEL N, HORVITZ Z, LEVIN R, et al. Figure seer:parsing result-figures in research papers[C]//Proceedings of the 14th European conference on computer vision. Amsterdam:Springer, 2016:664-680.
[35] CUI L, WU Y, LIU J, et al. Template-based named entity recognition using BART[EB/OL].[2023-11-06]. https://arxiv.org/pdf/2106.01760.pdf.
[36] PETRONI F, ROCKTASCHEL T, LEWIS P, et al. Language models as knowledge bases[EB/OL].[2023-11-06]. https://arxiv. org/pdf/1909.01066.pdf.
[37] DAI Z, YANG Z, YANG Y, et al. Transformer-xl:attentive language models beyond a fixed-length context[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1901.02860.pdf.
[38] JIA S, SHI J, LI M. Chinese open relation extraction and knowledge base establishment[J]. ACM transactions on Asian and low-resource language information processing, 2018, 17(3):15.1-15.22.
[39] 秦兵, 刘安安, 刘挺. 无指导的中文开放式实体关系抽取[J]. 计算机研究与发展, 2015, 52(5):1029-1035. (QIN B, LIU A A, LIU T. Unsupervised Chinese open entity relation extraction[J]. Journal of computer research and development, 2015, 52(5):1029-1035.)
[40] QIU L, ZHANG Y. ZORE:a syntax-based system for Chinese open relation extraction[C]//Proceedings of the 2014 conference on empirical methods in natural language processing. Doha:Association for Computational Linguistics, 2014:1870-1880.
[41] FADER A, SODERLAND S, ETZIONI O. Identifying relations for open information extraction[C]//Proceedings of the 2011 conference on empirical methods in natural language processing. Edinburgh:Association for Computational Linguistics, 2011:1535-1545.
[42] LOSHCHILOV I, HUTTER F. Decoupled weight decay regularization[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1711.05101.pdf.
[43] GAO J, LI X, XU Y, et al. Efficient knowledge graph accuracy evaluation[C]//Proceedings of the 45th international conference on VLDB. Los Angeles:Curran Associates, 2019:1679-1691.
[44] 李冰. 基于多领域本体的探索式搜索的查询推荐方法研究[D]. 沈阳:东北大学, 2014. (LI B. Study on exploratory search query recommendation method based on multi-domain ontology[D]. Shenyang:Northeastern University, 2014.)
[45] NEO4J. Neo4J desktop installer[EB/OL].[2023-11-01]. https://neo4j.com/download-thanks-desktop/?edition=desktop&flavour=winstall64&release=1.4.1&offline=true.
[46] 赵海平, 付婷, 刘伟超. 不同任务情境下研究生网络信息搜寻行为特征与差异研究[J]. 数字图书馆论坛, 2019(10):64-72. (ZHAO H P, FU T, LIU W C. Research on characteristics and differences of postgraduate online information seeking behavior under different task contexts[J]. Digital library forum, 2019(10):64-72.)
[47] 袁红, 周浩. 用户参与量表运用于探索式搜索效果评价的适用性探讨[J]. 情报理论与实践, 2020, 43(6):109-114. (YUAN H, ZHOU H. Applicability of user engagement scale in exploratory search effectiveness evaluation[J]. Information studies:theory & application, 2020, 43(6):109-114.)
[46] ZHONG Z, CHEN D. A frustratingly easy approach for entity and relation extraction[EB/OL].[2023-11-01]. https://arxiv.org/pdf/1711.05101.pdf.
Options
文章导航

/

〈 〉