Research On Express Service Quality Evaluation Based On Text Opinion Mining

With the change of express consumers' demand from basic service to high-end fine service,the express service market has changed from price competition to service quality brand conpetitioa The quality of express service has become the bottleneck for the survival and development of express enterprises and the improvement of their conpetitiveness.However,the way of using questionnaire to obtain customer satisfaction and evaluate service quality cannot meet the needs of express delivery enterprises to grasp customer needs,monitor and improve the quality of express delivery service.This paper proposes a method of evaluating express service quality based on text opinion mining,using express network comments as data source,in order to meet the needs of egress service quality evaluation and monitoring analysis.Firstly,this paper combs and summarizes the previous research on express service quality.Taking SERVQUAL model and LSQ model as prototypes,combining with the process and characteristics of express service,the dimensions are adjusted.The evaluation index system of oxpress service quality is preliminarily constructed,which includes 23 indicators of 7 dimensions,including the whole enterprise,charging,timeliness,courier,informationization,security and customer service.In order to get the factors of express service that consumers actually pay attention to,the data are investigated and analyzed,and the evaluation index system of express service quality is finally determined.Secondly,according to the fine-grained opinion mining method based on dependency parsing,which is used in express service comment mining,the operation of clause segmentation,part-of-speech tagging and dependency parsing are determined.Based on the analysis of the syntactic structure and semantic features of express service reviews,the extraction rules of express service evaluation units are summarized and designed.Based on the constructed emotional dictionary,and with the help of semantic similarity,anemotionalpolarity judgment algorithm and emotional intensity calculation rules are designed to complete the measurement of express service quality evaluation indicators.Then,on the basis of the express service quality evaluation index system and the index measurement scheme based on the express comment design,the evaluation object index classification algorithm based on prior knowledge and word2vec semantic similarity calculation and the calculation method of the emotional score of the secondary index are designed,and the processing of the index measurement is completed.The AHP judgment matrix is constructed by the number of consumer evaluation unite associated with the evaluation indicators,and the weight of indicators at all levels is determined.The measurement method of express service quality is designed,and the overall construction of express service quality evaluation model based on text opinion mining is completed.Finally,using the evaluation model of express service quality based on text opinion mining constructed in this paper,taking four links and one up,SF and EMS as examples,the empirical research and evaluation results are analyzed,and the validity and applicability of the model are demonstrated,which provides a basis for express enterprises to improve service quality and formulate solutions.In addition,it has certain guiding significance for consumers to choose express delivery enterprises reasonably.There are 41 figures,25 tables and 78 references in this paper.