摘要
针对目前领域知识图谱实体关系抽取效果不佳的问题,提出一种面向领域知识图谱的实体关系抽取模型研究方法.先建立由编解码模块、实体识别模块和实体关系抽取模块组成的实体关系抽取模型,在实体关系抽取模型中,通过双向长短期记忆神经网络对文本句子进行编码处理,将编码后文本句子特征表示向量输入至基于深度神经网络的实体识别模块中进行文本句子的实体识别,并将识别结果输入至基于卷积神经网络的实体关系抽取模块中进行实体关系抽取,然后将实体关系抽取获取的实体关系三元组输入至编解码模块中进行解码操作,实现最终的面向领域知识图谱的实体关系抽取.实验结果表明,该方法的实体关系抽取效果和整体应用效果较好.
Aiming at the problem of poor performance of entity relationship extraction in current domain knowledge graphs,we proposed a research method for entity relationship extraction models oriented towards domain knowledge graphs.Firstly,we established an entity relationship extraction model consisting of an encoding and decoding module,an entity recognition module,and an entity relationship extraction module.In the entity relationship extraction model,a bidirectional long short-term memory neural network was used to encode text sentences,and the feature representation vectors of the encoded text sentences were input into a deep neural network-based entity recognition module for entity recognition of text sentences,and the recognition results were input into the entity relationship extraction module based on convolutional neural networks for entity relationship extraction.Secondly,the entity relationship triplet obtained from entity relationship extraction was input into the encoding and decoding module for decoding operation,achieving the final entity relationship extraction for domain oriented knowledge graph.The experimental results show that the proposed method has better entity relationship extraction effect and overall application effect.
作者
何山
肖晰
张嘉玲
HE Shan;XIAO Xi;ZHANG Jialing(School of Computer Science and Software Engineering,Southwest Petroleum University,Chengdu 610599,China)
出处
《吉林大学学报(理学版)》
北大核心
2025年第2期465-471,共7页
Journal of Jilin University:Science Edition
基金
国家自然科学基金面上项目(批准号:62276099)。
关键词
知识图谱
实体关系抽取
实体识别
卷积神经网络
knowledge graph
entity relationship extraction
entity recognition
convolutional neural network
