Deep Learning Based B2C E-commerce Logistics Service Quality Evaluation

The rapid development of the Internet industry has greatly promoted social and economic progress,people's living standards have gradually improved,the amount of information on the online platform has exploded,and shopping methods are not limited to offline shopping.Online shopping has gradually become popular.It has become an indispensable part of people's lives.The changes in people's shopping habits have made the competition in the B2 C e-commerce industry increasingly greater.Low prices and high quality of goods have long been the only criteria for customers to choose online shopping platforms.Logistics is used as online shopping an important part of the process,the level of its service quality has become one of the important influencing factors for customers to choose online shopping platforms.At present,most of the scholars' research on the quality of B2 C e-commerce logistics services still stays in the traditional research methods,and there is no research method suitable for massive data.Based on this background,this paper aims at the quality of B2 C e-commerce logistics services through big data research methods.The following research has been carried out in order to help B2 C e-commerce logistics companies improve the quality of logistics services and enhance the competitiveness of the industry.(1)Through literature review and reference to LSQ model and SERVQUAL model,combined with the characteristics of B2 C e-commerce logistics services,based on the construction principles of B2 C e-commerce logistics service quality evaluation indicators,six B2 C e-commerce logistics service quality evaluations were constructed The dimensions are: economy,empathy,reliability,convenience,timeliness and interactivity.(2)Use the text preprocessing technology to clean the comment text data,segment the words,remove stop words,word frequency statistics,etc.,use the word frequency statistics and the Word2 vec model to build a B2 C e-commerce logistics service keyword library,and propose a screening logistics based on the keyword library Relevant useful text method,according to the definition of six dimensions of B2 C e-commerce logistics service quality evaluation,the logistics service keywords are classified,and the useful text is allocated to the six evaluation dimensions by matching the keywords with the review data.(3)Construct a deep learning-based B2 C e-commerce logistics service quality evaluation method.The result of this method is the sentiment multi-classification of logistics review data.The comment data is vectorized according to dimensions as input data of the deep learning model.The LSTM neural network model is used to extract the data features and perform sentiment classification.Finally,the sentiment classification label of the review data is obtained.(4)Finally,the constructed B2 C e-commerce logistics service quality evaluation method was used to conduct empirical research on several categories of product review data such as refrigerators,TVs,air conditioners,and washing machines in JD Mall.According to the research results,the problems in the logistics service process of JD Mall are found,and specific suggestions are provided for each dimension to provide a reference for enterprises to improve the quality of logistics services.