基于产品评论数据的KANO模型分析方法研究

Research on KANO Model Analysis Method Based on Product Review Data

产品评论中包含产品特征描述信息,客户对不同产品特征的偏好和情感倾向性不同,如何从产品评论中抽取产品特征描述,并从客户满意度角度评估产品特征的重要度对企业的产品改进决策有重要意义。利用LDA模型进行产品特征词提取,并结合BERT模型对产品评论进行文本特征分类和情感分类分析。根据产品评论信息,提炼出产品特征占比(PCFi)和产品特征重要度(PCDi)指标,在产品特征的情感分布基础上,提出了产品特征KANO分析方法并用空调产品评论数据进行了实证研究。通过利用产品评论数据对产品特征进行KANO分析能够为企业进行产品改进提供有价值的参考信息。

Product reviews contain product feature description information,customers have different preferences and emotional tendencies for different product features,how to extract product feature descriptions from product reviews and to evaluate the importance of product features from the perspective of customer satisfaction is of great significance to the company's product improvement decisions.Use the LDA model to extract product feature words,and combine the BERT model to perform text feature classification and sentiment classification analysis on product reviews.According to the product review information,the product feature percentage(PCFi)and product feature importance(PCDi)indicators were extracted.Based on the emotional distribution of product features,the product feature KANO analysis method was proposed and the air conditioning product review data was used for empirical research.KANO analysis of product characteristics by using product review data can provide valuable reference information for companies to improve their products.