考虑特征差异的地铁乘客满意度研究

A Study of Metro Passenger Satisfaction Considering Differences in Characteristics

摘要: 基于武汉地铁乘客满意度调查中获取到的基本特征，采用K-means++算法识别出外地乘客、男性通勤、女性通勤以及生活休闲四类群体，并对各群体主要特征进行分析。依据满意度调查中各项指标收集到的具体问题，利用LDA主题模型识别出其主题分布与主题词，明确各类群体的诉求，并提出了相应的建议。研究结果可为武汉地铁提升服务质量提供参考和依据。

Abstract: Based on the basic characteristics obtained from the Wuhan Metro passenger satisfaction survey, the K-means++ algorithm is used to identify four groups of foreign passengers, male commuters, female commuters and life and leisure, and the main characteristics of each group are analyzed. According to the specific problems collected from each index in the satisfaction survey, the LDA topic model is used to identify their topic distribution and subject terms, clarify the demands of each group, and put forward corresponding suggestions. The research results can provide refer-ence and basis for Wuhan Metro to improve its service quality.