基于消极语义网络的产品改进策略探索  

Research on Product Improvement Strategies Based on Negative Semantic Networks

作　　者：王璐 张孟珂 WANG Lu;ZHANG Mengke(Economic Investigation Department,Henan Police College,Zhengzhou 450046,China;Tianjin University of Technology,Management College,Tianjin 300100,China)

机构地区：[1]河南警察学院经济犯罪侦查系,郑州450046 [2]天津理工大学管理学院,天津300100

出　　处：《安阳工学院学报》2025年第2期72-76,共5页Journal of Anyang Institute of Technology

基　　金：河南省科技攻关项目(252102320052);河南省高校人文社会科学研究一般项目(673);河南警察学院科研项目(HNJY-2024-SSZX-47);河南省社科联调研课题(SKL-2024-650)。

摘　　要：情感分析结合语义网络的协同分析,可以实现多企业、多产品的纵向与横向对比分析,提升情感分析的维度。通过Rython网络爬虫技术获取华为、小米等65款手机的京东线上评论数据,用jieba分词处理,然后采用SnowNLP(Natural Language Processing)情感分析模型对多款手机评论进行情感评分,选取消极评论使用复杂网络分析模型Force Atlas2进行语义网络建模,以网络形式呈现关键词间的内在关联。结果表明:品牌方面,华为、荣耀相比其他品牌消极评论较多;性能方面,拍照效果、系统流畅度等有待提升;服务方面,物流、包装等环节需要改进。这些用户反馈的智能手机使用痛点,能够为企业明确产品性能的优化方向提供参考。The collaborative analysis of sentiment analysis and semantic network analysis can realize the vertical and horizontal comparative analysis of multiple enterprises and their products,thus enhancing the dimension of sentiment analysis.In the experiment,the SnowNLP sentiment analysis model is used to score the sentiment of comments on several mobile phones.The negative comments are selected and the ForceAtlas2 complex network analysis model is used for semantic network modeling to present the internal relationship between key words in the form of a network.The research results show that in terms of brands,Huawei and Honor have more negative comments than other brands.In terms of performance,the photo-taking effect,system fluency,etc.need to be improved.In terms of service,the logistics,packaging and other links need to be improved.These pain points of smartphone use feedback by users can provide enterprises with a reference for clarifying the optimization direction of product performance.

关 键 词：文本挖掘 情感分析 语义网络 产品竞争力 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]