Defect Identification Of Express Service Based On Customer Online Reviews

In recent years,with the continuous development of e-commerce,express service has become an essential service in public life,and also plays an important role in the social and economic development.For enterprises,the quality of service has a direct impact on the business performance of express delivery enterprises.If there are defects in the service,the market competitiveness of enterprises will be greatly weakened.At present,the overall development of express service industry is unbalanced and inadequate.Therefore,it is necessary to reform and innovate the express service and further improve the quality of express service.Some researches have designed survey questionnaire to provide service quality improvement strategies for express delivery enterprises.However,the number of questionnaires collected is limited and the timeliness is insufficient,which can not play a role in real-time monitoring of service quality for enterprises.With the continuous development of social media,online reviews has become a new carrier of customer feedback.Mining valuable information from it is of great significance for enterprises to find defects in express service and improve the quality of express service.Therefore,this paper takes online reviews as the data source,directly mining service defects from them,and combining with the traditional model for analysis,exploring service quality management from the perspective of customers.In this paper,we use Python to compile our own crawler program,and take the post content of STO express and YTO express in Baidu Tieba for five years,that is,online reviews,as the data source.First of all,the text analysis framework is constructed,which includes four types of features to screen the comments about the defects of express service.Using text classification algorithm to predict,get defect comment set,provide more accurate data for service defect identification;secondly,using LDA(Latent Dirichlet Allocation,LDA)model,getting defect feature words,and summarizing ten defect attributes;then,using fine-grained sentiment analysis,based on sentiment dictionary,calculate ten defect attributes respectively.At the same time,the importance of each defect attribute is evaluated by statistics of the proportion of comments related to each defect attribute.Finally,using the importance and defect degree,the Importance Defect Degree Competitor Analysis(IDCA)model is constructed to form four quadrants of different nature,and corresponding service defect coping strategies are provided.Ten kinds of service defects are mapped in four quadrants,and classified and analyzed qualitatively.At the same time,opportunity algorithm is used to quantify the improvement degree of defect attributes in each quadrant,to provide priority ranking,and to comprehensively formulate response plans.In addition,through the independent analysis of the annual data,forming a three-dimensional time series analysis chart,mining the dynamic changes and trends of various types of express service defects,and providing supplementary suggestions for the formulation of response strategies for express service defects.This paper uses text mining and traditional model to explore service quality defects,which has a certain theoretical and practical contribution to service quality management.At the theoretical level,the paper introduces the traditional IPA(Importance-Performance Analysis,IPA)into the field of service defect management,and constructs an applicable IDCA model,which provides a new idea for the analysis and response of service defects.At the same time,it combines traditional model and algorithm with text mining,enriches the deduction of them.At the practical level,this study provides a complete service defect identification framework for express companies,which can help managers identify service defects automatically and quickly in real time,and provide managers with scientific response strategies through classification and priority sorting.