摘要
针对航空装备使用和保障人员获取信息效率低的问题,将文本相似度匹配技术引入航空装备维修保障问答领域,结合规则模板、生成模型等方法生成相似问句,提出一个航空装备维修保障领域问句相似度数据集;融入外部信息知识,通过领域词典识别问句对中的关键词,对CoSENT(Cosine Sentence)模型引入关键词注意力机制和特征交互机制提升模型性能。在自建数据集和讯飞文本相似度数据集上进行验证,与基准模型CoSENT相比,论文方法的Spearman相关系数分别提升了2.3%和1.7%。
Aiming at the problem of low efficiency in obtaining information for aviation equipment users and support personnel,the text similarity matching technology is introduced into the field of aviation equipment maintenance support question answering.Combining with the rule template and generation model to generate similar questions,a similarity data set of questions in aviation equipment maintenance field is proposed.External information knowledge is integrated into model,keywords in sentence pairs are identified through domain dictionaries.In addition,keyword attention mechanism and feature interaction mechanism are introduced into the model to improve the performance of CoSENT(Cosine Sentence).It is verified on the self-built data set and IFLYTEK text similarity data set.Compared with the benchmark model CoSENT,the Spearman correlation coefficient of this method is increased by 2.3% and 1.7% respectively.
作者
翟一琛
顾佼佼
刘涛
姜文志
ZHAI Yichen;GU Jiaojiao;LIU Tao;JIANG Wenzhi(Naval Aviation University,Yantai 264001)
出处
《舰船电子工程》
2023年第9期145-150,共6页
Ship Electronic Engineering
