Volume 2,Issue 7
一种基于知识图谱和联合自然语言模型的知识问答方法
针对智能问答系统中用户意图识别准确率不足及旅游领域知识图谱资源匮乏的问题，提出了一种基于大连市红色旅游知识图谱与ERNIE-BILSTM-CRF联合模型的知识问答系统。系统构建了涵盖城市、地区、景区、旧址、背景、意义及门票价格七类节点的红色旅游知识图谱，利用Neo4j图数据库实现数据的结构化存储与可视化查询。为提升语义理解能力，创新性地将百度预训练语言模型ERNIE与双向长短时记忆网络（BILSTM）、条件随机场（CRF）相结合，构建联合模型用于命名实体识别，显著提高了关键实体和意图的识别准确率。实验基于5000条大连红色旅游问答数据集，结果显示ERNIE-BILSTM-CRF模型在精确率、召回率及F1值上分别达到98.99%、99.75%和98.91%，优于多种对比模型，验证了模型的有效性和鲁棒性。节点类型测试进一步表明系统对不同类别实体均具备较高识别能力。本文丰富了旅游领域知识图谱资源，并为智能问答系统的自然语言理解与知识检索提供了新思路，推动了旅游信息智能化服务。
[1]Brown T, Mann B, Ryder N, et al. Language models are few-shot learners[J]. Advances in neural information processing systems, 2020, 33: 1877- 1901.
[2]Ouyan L, Wu J, Jiang X, et al. Training language models to follow instructions with human feedback [J]. Advances in neural information processing systems, 2022, 35: 27730-27744.
[3]Etzioni O, Cafarella M, Downey D, et al. Unsupervised named-entity extraction from the web: An experimental study[J]. Artificial intelligence, Elsevier, 2005, 165(01): 91–134.
[4] 王荣坤.基于旅游知识图谱的智能问答系统设计与实现[D].青岛大学,2022.DOI:10.27262/d. cnki. gqdau.2022.001953.
[5]WANG L, MA C, FENG X, et al. A survey on large language model based autonomous agents [J]. arXiv preprint arXiv:230811432, 2023.
[6]Auer S, Bizer C, Kobilarov G, et al. Dbpedia: A nucleus for a web of open data[M]//The semantic web. Springer, Berlin, Heidelberg, 2007: 722-735.
[7]Suchanek F M, Kasneci G, Weikum G. Yago: a core of semantic knowledge[C]//Proceedings of the 16th international conference on World Wide Web. 2007: 697-706.
[8]Niu X, Sun X, Wang H, et al. Zhishi.me-weaving chinese linking open data[C]//International Semantic Web Conference. Springer, Berlin, Heidelberg, 2011: 205-220.
[9]Wang H, Wu T, Qi G, et al. On Publishing Chinese Linked Open Schema[M]//The Semantic Web-ISWC 2014. Springer International Publishing, 2014: 293-308.
[10]Wang Z, Li J, Wang Z, et al. XLore: A Large-scale English-Chinese Bilingual Knowledge Graph[C]// International semantic web conference (Posters & Demos). 2013,1035: 121-124.
[11]IMDB Official. IMDB[EB/OL]. [2020-11-23]. http: //www.Imda.com．
[12]BAI T, GONG L, ANG Y, et al. A method for exploring implicit concept relatedness in biomedical knowledge network[J]. BMC Bioinformatics, 2016, 17(9): 53-66．
[13]ROSPOCHE R M, VAN E R P M, VOSSEN P, et al. Building event-centric knowledge graphs from news[J]. Journal of Web Semantics, 2016, 37: 132-151．
[14]Aliod D M, van Zaanen M, Smith D. Named entity recognition for question answering[C]. Australasian Language Technology Workshop (ALTA). Australia and New Zealand: ALTA, 2006: 51-58.
[15]SHI L, LI S, YANG X, et al. Semantic health knowledge graph: Semantic integration of heterogeneous medical knowledge and services[J]. BioMed Research International,2017,1(4): 1-12.
[16]Cheng P, Erk K. Attending to entities for better text understanding[C]. Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Menlo Park, CA, USA: AAAI, 2020, 34(05): 7554-7561.
[17]Bordes A, Usunier N, Chopra S, et al. Large-scale simple question answering with memory networks[J]. arXiv preprint arXiv: 1506.02075, 2015.
[18] 王子岳.基于深度学习的自然语言理解 模型[D].南京邮电大学,2021.DOI:10.27251/ d.cnki.gnjdc. 2021.000042.
[19] 王博,群诺,王京博,等.基于知识图谱的西藏旅游问答系统设计与实现[J].信息技术与信息化,2023,(12):181-184.
[20] 鄢晗晖.基于知识图谱的问答系统的设计与实现[D].华东师范大学,2022. DOI:10.27149/d. cnki.ghdsu.2022.000894.
[21]SHANAHAN M. Talking about large language models[J]. Communications of the ACM, 2024, 67(2): 68-79.
[22]TAYLOR R, KARDAS M, CUCURULL G, et al. Galactica: A large language model for science[J]. arXiv preprint arXiv:221109085, 2022.
[23] 罗琳凡.基于预训练模型的旅游领域知识图谱构建及智能问答应用[D].东南大学,2022.DOI:10.27014/d.cnki.gdnau.2022.001730.
[24] 张峻菖.基于ERNIE-BiGRU-CRF的山西旅游领域命名实体识别研究[D].山西财经大学,2024.
[25] 徐捷,邵玉斌,杜庆治,等.结合混合特征提取与深度学习的长文本语义相似度计算[J].计算机工程与科学,2024,46(08):1513-1520.