论文部分内容阅读
不同性别的用户对产品的看法与品位存在着差异,特别是在欣赏与时尚相关的产品上,性别对用户判断的影响显得尤为重要。根据电子商务中在线商品的浏览记录,采用支持向量机(support vector machines,SVM)对所选取的7个特征建立模型,并进行性别判断。经过模型分析和训练,准确率可达79.21%。同时讨论了网络购物与实体店购物的区别,并对SVM进行了核函数对比及其它性能的研究,从理论和实际应用上为核函数的选取和SVM的选用提供参考。
Different gender users have different views on the product and taste, especially in fashion-related products, the impact of gender on the user’s judgments is particularly important. According to the browsing records of online goods in e-commerce, we use the support vector machines (SVM) to establish the model of the selected seven features and judge the gender. After model analysis and training, the accuracy rate can reach 79.21%. At the same time, the difference between online shopping and physical shopping is discussed. The comparison of the kernel function and other properties of the SVM is also discussed. The selection of the kernel function and the selection of SVM are provided for the theoretical and practical applications.