摘要
为有效处理文本文献多语种检索词翻译歧义、词意模糊等问题,提出基于孪生神经网络的文本文献多语种混合检索算法。利用孪生神经网络提取文本文献词汇特征,创建可微分代价函数,将代入空间变换器网络作为孪生神经网络前级网络,以此提升多语种特征提取准确性。构建链接模型,利用贝叶斯定理计算,根据数值结果利用文献相关性DocRank算法对多语种文本文献进行预处理,通过优先度分数获取精准的多语种混合检索结果。实验结果表明,所提方法运行算耗时短、检索精度高,具有较强实用性。
In order to effectively deal with the problems of translation ambiguity and ambiguity of text documents in multilingual search terms,a text document multilingual hybrid retrieval algorithm based on twin neural network is proposed.The twin neural network is used to extract the vocabulary features of text documents,create a differentiable cost function,and substitute the space transformer network as the pre-stage network of the twin neural network to improve the accuracy of multilingual feature extraction.It constructs a link model,uses Bayes'theorem to calculate,use the document relevance DocRank algorithm to preprocess multilingual text documents based on the numerical results,and obtains accurate multilingual mixed retrieval results through priority scores.The experimental results show that the proposed method has a short running time,high retrieval accuracy,and strong practicability.
作者
林宗英
林民山
LIN Zong-ying;LIN Min-shan(Institute of Intelligent Manufacturing,Quanzhou Vocational and Technical University,Quanzhou 362000 China)
出处
《自动化技术与应用》
2023年第9期103-106,共4页
Techniques of Automation and Applications
基金
福建省中青年教师教育科研项目(科技类)(JAT201192)。
