Word2Vec-KNN技术支持下潮流玩具质量检测模型研究  

Research on quality inspection model for trendy toys supported by Word2Vec-KNN technology

作　　者：吕远智 LV Yuanzhi(Beijing Bubble Mart Cultural and Creative Co.,Ltd.,Beijing 100123,China)

机构地区：[1]北京泡泡玛特文化创意有限公司,北京100123

出　　处：《计算机应用文摘》2024年第10期92-94,共3页Chinese Journal of Computer Application

摘　　要：随着人们生活水平的提高,越来越多的消费者更加注重所购产品的质量,特别是在儿童玩具方面。质量不合格的玩具产品会给儿童带来诸多影响,包括但不限于安全隐患及对儿童健康产生的影响。然而,工业制造中的产品质量检测报告种类繁多且不易被理解,无法直观体现产品质量。因此,文章提出了一种基于Word2Vec与K最近邻分类算法相结合的产品质量评估模型。该模型能够通过产品质量报告对某玩具进行评估,从而判断其质量。实验结果表明,在数据集尺寸达到900时,K均值聚类算法模型、局部加权最近邻算法模型和混合模型算法模型的准确率分别为0.84,0.91与0.96,损失函数值分别为0.07,0.05及0.03,证明所提模型能够对玩具产品进行准确评估,从而为消费者和质量监管部门提供一定的决策支持。With the improvement of people̓s living standards,more and more consumers are paying more attention to the quality of their purchased products,especially in the field of children̓s toys.Toy products with substandard quality can have many impacts on children,including but not limited to safety hazards and their impact on children's health.However,there are many types of product quality inspection reports in industrial manufacturing that are difficult to understand and cannot intuitively reflect product quality.Therefore,the article proposes a product quality evaluation model based on the combination of Word2Vec and K-nearest neighbor classification algorithm.This model can evaluate a toy through a product quality report to determine its quality.The experimental results show that when the dataset size reaches 900,the accuracy of the K-means clustering algorithm model,local weighted nearest neighbor algorithm model,and hybrid model algorithm model are 0.84,0.91,and 0.96,respectively,and the loss function values are 0.07,0.05,and 0.03,respectively.This proves that the proposed model can accurately evaluate toy products and provide certain decision support for consumers and quality supervision departments.

关 键 词：产品质量评估 K最近邻 Word2Vec 大数据 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]