Analysis On Regional Difference Of Logistics Service Level Based On Multi-source Data

With the rapid development of the Internet and the expansion of online shopping,China’s express industry has also expanded rapidly.The China Express Development Index shows that China’s Express Industry Development Scale Index in 2018 was 1765.3,an increase of25.5 points over 2017.And Service Quality Index It is 133.4,a year-on-year increase of30.2%.This is the first time that the growth rate of the Service Index exceeds the Development Scale Index,which means that after the large-scale and high-speed growth of the logistics industry in China,it is gradually changing from a large logistics country to a logistics power country.And improving the quality of services become a key part of service management in the logistics industry.At present,researchers at home and abroad have discovered the value of user behavior data on the Internet,and various types of research using user comment are gradually developing into one of the important areas of big data analysis.In this context,this article chooses to use a variety of Centralized logistics evaluation information,that is,the State Post Office logistics complaint case,Weibo online reviews,and T-mall product reviews as data sources,analyze and study the service quality of the logistics industry through a weakly supervised LDA model.First,use Python to target the data source perform data collection,text preprocessing,word segmentation and other processing to form a usable data set.Extract features based on syntactic analysis,Word2 vec word vector training,and K-means clustering for feature extraction and cluster analysis of the data.Second,combined with the correction The SERVQUAL and LSQ evaluation models establish the six dimensions of logistics service level evaluation,that is,the six themes of the topic model.Based on prior knowledge in the logistics field,the seed words are manually generated based on the clustered logistic service feature set.According to the generated seeds Words,using a weakly supervised LDA model to extract the underlying subject structure of the logistics service quality from the dataset,That is,the six dimensions of the service level evaluation model.Use statistical methods to classify and analyze topics,and obtain the impact of each service quality dimension on logistics service evaluation for different research areas.Finally,based on the logistics service levels in different regions Differences,and put forward the path to improve the service quality of logistics enterprises in response to the outstanding problems shown in the analysis results,with a view to improving the service quality level of the logistics industry.The study found that there is no significant difference in the reliability of logistics service levels in different regions,but the regional differences in the dimensions of convenience,safety,and economy are large.Among them,the southwest and northwest regions are more dissatisfied with convenience,and the economic impact on the evaluation of logistics service levels in the central and western regions is significantly greater than in the southeast.And Beijing,Shanghai and other places have lower dissatisfaction factors,but higher interactive dissatisfaction.The research conclusions of this paper have certain practical value for logistics companies to optimize their operating strategies and improve logistics services,and provide a reference for detailed research on logistics service levels in different regions.