短生命周期类电子产品的用户需求挖掘研究  

Research on the User Demand Mining of Short Life Cycle Electronic Products

作　　者：张亚菲 申彦[1] 敬露艺 ZHANG Ya-fei;SHEN Yan;JING Lu-yi(School of Management,Jiangsu University,Zhenjiang 212013,China)

机构地区：[1]江苏大学管理学院,江苏镇江212013

出　　处：《软件导刊》2021年第9期217-222,共6页Software Guide

基　　金：国家自然科学基金项目(61702229,71573107);全国统计科学研究计划项目(2016LY17)。

摘　　要：为快速、准确地挖掘短生命周期类电子产品快速转变的用户需求,研究提出一种利用用户生成内容(UGC)对短生命周期类电子产品进行用户需求挖掘的框架。首先,分别利用LDA和SnowNLP提取UGC特征主题并对其进行情感极性分类;然后,借助多元逻辑回归模型得出各特征主题的不同情感倾向对评分的影响;最后,利用Kano模型实现对用户需求的分类与排序。利用该框架分析了电商平台上有关智能终端的UGC,成功获取了用户不同种类的需求。实验结果表明,所构建的框架准确率达到82%以上。该框架可以有效地从UGC中识别产品特征主题、划分情感类别,从而快速、准确地挖掘出短生命周期类电子产品用户不同类别的需求,为企业相关部门制定战略决策提供依据。In order to tap the rapidly changing user needs of short-life-cycle electronic products quickly and accurately,a framework for user demand mining of short-life-cycle electronic products by using UGC is proposed.First,LDA and SnowNLP are used to extract the characteristic themes of UGC and classify them into affective polarities.Then,using multiple logistics regression model,the influence of different emotional tendency of each characteristic topic on the scoring result is obtained.Finally,a Kano model is used to realize the classification and ordering of user requirements of short life cycle electronic products.The framework is used to analyze UGC of intelligent terminals on an e-commerce platform,and the different demands of users are obtained successfully.The experimental results show that the accuracy of the framework is more than 82%.The constructed framework can identify product feature themes and classify emotion categories from UGC effectively,so as to excavate the needs of different kinds of users of short-life-cycle electronic products quickly and accurately and provide decision support for the formulation of strategies of relevant departments of enterprises.

关 键 词：用户需求 UGC 短生命周期类电子产品 特征提取 情感分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]