基于改进最小二乘支持向量机的电商细粒度数据采集检测研究  

Research on fine-grained e-commerce data collection and detection based on improved least squares support vector machine

作　　者：郭戈 GUO Ge(Yangquan Vocational And Technical College,Yangquan Shanxi 045000,China)

机构地区：[1]山西阳泉职业技术学院,山西阳泉045000

出　　处：《自动化与仪器仪表》2024年第10期69-74,共6页Automation & Instrumentation

摘　　要：针对电商细粒度数据采集和检测不足,以及电商细粒度目标评论与产品评论文本语义不足的情况,研究在蚁群算法和支持向量机算法的基础上提出了新的蚁群——动态聚类-最小二乘支持向量机算法,新算法将电商细粒度数据进行分析检测,进一步提升商家对数据的分析研究。研究结果表明,SD-LS-SVM算法模型比其他三种算法模型准确率分别高17%、10%和7%,SD-LS-SVM算法模型在处理电商细粒度数据时表现优异,具有高准确率和稀疏度。与其他算法相比,它在情感倾向数据处理中获得最高准确率92%。此外,数据分析能显著提升电商产品销售量,而SD-LS-SVM算法模型在销售额变化中表现最佳,明显优于其他算法模型,具有实用性和改善电商产品销售和营业的效果。最后研究搭建的系统能够有效对电商数据进行分析,并提供更加直观的细粒度分析数据。由此可见新算法模型对电商细粒度数据的分析检测研究有一定的贡献,实现了对电商产品细粒度的意见挖掘与数据分析。In response to the insufficient collection and detection of fine-grained data in e-commerce,as well as the insufficient semantic meaning of fine-grained target comments and product comment texts in e-commerce,a new ant colony dynamic clustering least squares support vector machine algorithm is proposed based on ant colony algorithm and support vector machine algorithm.The new algorithm analyzes and detects fine-grained e-commerce data,further improving the analysis and research of merchants on data.The research results show that the SD-LS-SVM algorithm model has an accuracy rate 17%,10%,and 7%higher than the other three algorithm models,respectively.The SD-LS-SVM algorithm model performs well in processing fine-grained e-commerce data,with high accuracy and sparsity.Compared with other algorithms,it achieved the highest accuracy of 92%in sentiment orientation data processing.In addition,data analysis can significantly improve the sales volume of e-commerce products,and the SD-LS-SVM algorithm model performs the best in sales changes,significantly better than other algorithm models,with practicality and the effect of improving e-commerce product sales and business.Finally,the system constructed can effectively analyze e-commerce data and provide more intuitive fine-grained analysis data.It can be seen that the new algorithm model has made certain contributions to the analysis and detection of fine-grained data in e-commerce,achieving opinion mining and data analysis of fine-grained e-commerce products.

关 键 词：电商细粒度数据 SD-LS-SVM算法 产品销量 分析检测 准确率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]