Research On Service Quality Management Of Online Ride-hailing In The Sharing Economy Based On PCN

Supported by the rapidly development of mobile internet,business models based on sharing economy are growing at a spectacular rate and permeate through people's daily lives.As a totally new type of business model,the growing sharing economy is modifying consumer behavior and furthermore,makes a significant impact on the original industrial structures,employment modes and social relationships.Among multiple cases of sharing economy,online ride-hailing has successfully practiced the idea of sharing in transportation field,resulting in the emergence of large-scale sharing platforms,including Uber and Didi Chuxing,which have acquired large numbers of users.However,although online ride-hailing is the typical representative of service industry,there still exits some problems in its service quality,as a result of its explosive growth.With the increasing service quality requirement from customers,sustainable improvements in service quality have become a key factor to the further development of online ride-hailing.Hence,this research takes Didi Chuxing as the study case and gathers data about its word-of-mouth from the internet.Then,existing issues with Didi Chuxing's online ride-hailing service is dug out though data mining technology,and its service network is analyzed with process chain network.Finally,optimized countermeasures for its service quality are proposed according to the analysis results of existing issues and service network.The study provide lessons for the establishment of data-driven operations and management.The content of the study is organized as follow:(1)Classification model based on LSTM for the internet word-of-mouth of online ride-hailing platform.Data about Didi Chuxing's internet word-of-mouth is gathered from Sina microblog with Python program and then,the collected data is preprocessed according to the filtering and cleaning rules.With the clean data,a text classification model based on LSTM is designed and trained,which can classify the internet word-of-mouth into passengers' comment,drivers' comment or marketing microblog automatically.It enables the platforms to analyze different types of data respectively and dig out valuable information.The input of the model is word embedding coming from the training results of CBOW(Continuous Bag-of-Words)model,with Chinese Wikipedia as training corpus.The classification accuracy of the model is 90.89%.(2)Service issues mining for online ride-hailing platform.The study makes a sentiment analysis for the passengers' comments and extract the negative comments out according to the analysis results.Then the task of service issues mining can be transformed into a task of frequent item-sets mining.Using FP-Growth algorithm,the study digs out the frequent co-occurrence words from the negative comments of passengers.By the analysis of the discovered co-occurrence words,the existing service issues are found.(3)Service quality management for online ride-hailing platforms based on PCN.To describe the service network of online ride-hailing,the study analyzes its service process and divides every process into multiple sub-steps.Then the PCN graph of online ride-hailing service can be drew according to which process region each sub-step belongs to.Combining the results of service issues mining,the study divides the issues into different types by their causes.Finally,optimized countermeasures for the service quality of online ride-hailing are proposed according to the corresponding optimization principles of PCN,with the aim of improving its customer satisfaction and promoting its healthy and sustainable development.