基于检索重排序模型的文本差异化研究

门业堃; 钱梦迪; 于钊; 滕景竹; 陈少坤; 颜旭

门业堃,钱梦迪,于钊,滕景竹,陈少坤,颜旭.基于检索重排序模型的文本差异化研究[J].电测与仪表,2023,60(1):57-63.
Men Yekun,Qian Mengdi,Yu Zhao,Teng Jingzhu,Chen Shaokun,Yan Xu.Text Differentiation Based on Reordering Retrieval Model[J].Electrical Measurement & Instrumentation,2023,60(1):57-63.

基于检索重排序模型的文本差异化研究

Text Differentiation Based on Reordering Retrieval Model

DOI：10.19753/j.issn1001-1390.2023.01.009

中文关键词: 差异性检索排序学习文本相似度检索重排序设备质量评估自动化

英文关键词:differential retrieval ranking learning text similarity retrieval reordering automation of equipment quality assessment

基金项目:

作者中文名	作者英文名	单位
门业堃	Men Yekun	国网北京市电力公司电力科学研究院
钱梦迪	Qian Mengdi	国网北京市电力公司电力科学研究院
于钊	Yu Zhao	国网北京市电力公司电力科学研究院
滕景竹	Teng Jingzhu	国网北京市电力公司电力科学研究院
陈少坤	Chen Shaokun	北京恒华龙信数据科技有限公司
颜旭	Yan Xu	北京恒华龙信数据科技有限公司

摘要点击次数: 1151

中文摘要:

电力行业在设备质量评估中需要结合行业标准规范文件中指定的具体标准来对设备质量进行准确评估。文中通过基于检索重排序模型的文本差异化模型,建立自动化、信息化、智能化的标准差异化梳理技术,有效解决了当前设备质量评估时的费时费力问题,并提升了文本差异检索的准确率。文中主要围绕自动化、信息化、智能化的标准差异化梳理技术,通过基于检索重排序模型的信息检索模型,建立同一领域的不同标准文本的检索比对,检索出不同文件对同一技术有着不同要求的差异性内容并预警提示。文中的创新点是,利用检索重排序精确性高的优点,在保留传统差异性检索召回模型的召回候选能力的基础上进一步提高精确性。模型在真实的电力行业技术标准文档上进行了系统鲁棒的交叉验证,验证了提出的模型效果优异,模型具有良好的实用性,可以广泛应用于电力设备质量评估,供应商评价标准检索等领域。

英文摘要:

The power industry needs to combine the specific standards specified in the industry standard specifications to accurately evaluate the quality of the equipment. Through the text differentiation model based on the retrieval and reordering model, this paper establishes standardization techniques for automation, informationization, and intelligent differentiation, which effectively solves the time-consuming and labor-intensive problems of current equipment quality assessment, and improves the accuracy of text difference retrieval . The article mainly focuses on the standardization and differentiation techniques of automation, informationization, and intelligence. Through the information retrieval model based on the retrieval reordering model, the retrieval comparison of different standard texts in the same field is established, and different documents retrieved have different requirements for the same technology. Different content and early warning tips. The innovation point in this paper is to further improve the accuracy on the basis of retaining the recall ability of the traditional differential retrieval and recall model by taking advantage of the high accuracy of retrieval reordering. The model was systematically and robustly cross-validated on the actual technical standards documents of the power industry, and it was verified that the proposed model has excellent results and the model has good practicability. It can be widely used in power equipment quality evaluation, supplier evaluation standard retrieval and other fields.

查看全文查看/发表评论下载PDF阅读器

关闭