关于深度学习的在线评论挖掘及需求获取方法  

On-line Comment Mining and Requirement Acquirement Method Based on Deep Learning

作　　者：李美[1] 裴卉宁 丁满[1] LI Mei;PEI Huining;DING Man(School of architecture and art design,Hebei University of technology,Tianjin 300401,China)

机构地区：[1]河北工业大学建筑与艺术设计学院,天津300401

出　　处：《机械设计与研究》2021年第6期184-189,共6页Machine Design And Research

基　　金：河北省社会科学基金项目(HB20YS046)。

摘　　要：针对传统用户需求获取方法数据量少、实时性差、成本过高等问题,提出一种基于深度学习获取用户生成内容(User-Generated Content,UGC)有效数据的在线评论挖掘及需求获取方法。以电子商务平台在线评论信息为数据源构建UCC在线评论语料集,并提出了一种结合机器学习与人工审查的需求获取方式,通过卷积神经网络过滤大型UGC在线评论语句的非信息性内容,并利用数据分析工具对用户需求嵌入密集的相似性语句进行聚类分析,以避免对重复性内容进行采样。最后由专业分析人员对机器学习所提取的用户需求进行有效审查以提升需求信息准确率。通过老年代步车设计需求获取过程为案例验证模型,证明了所研究方法的高效性和准确性。利用对用户反馈信息的分析,为企业新产品研发提供了理论支持与技术支撑。In order to solve the problem of small data size,poor real-time performance and high cost in the traditional method for oblaining user requirements,an online comment mining and requirements acquisition method based on User-Generated Content(UGC)is proposed.The online comment corpus of UGC is constructed based on the online comment information of e-commerce platform,and a requirement acquisition method combining machine learning and manual review is proposed.This method filters the non-informative content in large online comment statements by convolutional neural network,and uses data analysis tools to cluster the similar statements which are densely embedded in user requirements.Finally,prfesonal analysts review the user requirements extracled by machine learing method o improve the accuracy of requirements information.Tak ing the elderly vehicle design requirements acquisition process as a case,the proposed method proves fficient and accurate.Through the analysis of user feedback information,it can provide theoretical and technical support for the research and development of new products.

关 键 词：用户需求 用户生成内容 在线评论挖掘 深度学习:卷积神经网络 

分 类 号：TB472[一般工业技术—工业设计]