How to dig latent and useful logistics information resources from massive comment data of e-commerce logistics service is the foundation and key to improve logistics management and service quality. This paper aims to deal with text contents of E-commerce logistics service comments based on Chinese word segmentation, and uses LDA theme model to discover the potential topics in the comment texts of e-commerce logistics services, then applies the method of word frequency statistics to extract the key words of logistics service comment texts and makes its results visible, so as to accurately discover the key factors and information of logistics service comments. The paper uses the part of the logistics company's 3178 evaluation data on dianping.com, which automatically grabs through the web crawler technology as an example, and makes a comparative analysis of the text of logistics service comments on the dimension of company by using this text analysis method.