Research On The Construction Of E-commerce Platform And Sentiment Analysis

In recent years,with the improvement of national consumption level,the convenience of logistics and the technical support of the Internet field,many electronic business platforms have developed rapidly.Among them,there are Taobao,Tmall,Jingdong group Jingdong mall,vip.com attracting consumers with low-priced luxury goods,as well as NetEase Koala Purchase which focuses on cross-border electricity supplying.To adapt the larger amount of business,these electronic business platforms,on one hand,innovate independently on their own technical architecture,and form a set of stable system to support high concurrent business.On the other hand,these platforms analyze the data generated by users on the platform through data mining,artificial intelligence and other means to obtain the commercial value that hidden in the data.These data play a very important role in making recommendations for users and making operating strategies for merchants.This paper,aiming at making detailed design for every functional modules of e-commerce platform,is going to make a platform system,which is enough reliable and functional,by analyzing the platform's architecture,studying the technology selection and application of common business platform,for instance,load balancing,message queue,caching,data storage and search which are used in the electronic business platform.In addition,the text analysis algorithm of e-commerce review have also been studied,and the user evaluation and analysis module based on emotional dictionary is implemented.The main research contents of this paper include the following three aspects:(1)The in-depth study of the composition of electronic business platform and the link between it and modules.Through the demanding analysis,the business platform is divided into member,commodity,shop,order,shopping cart,payment interface and analysis of evaluation,which are core modules.Then these functions are extend to the detailed business requirements,and design the corresponding table structure and select the appropriate means of technical implementation.Among which,the evaluation and analysis module is mainly analyzing the commodity reviews.Through this module,the platform can obtain the users'emotional tendency to the.goods,so it will be regarded as an independent chapter to have key study.(2)The Implementation of an algorithm based on sentiment dictionary for users'evaluation and analysis.For a commodity comment sentence,word segmentation is the first thing to carry out,and then extract the word or phrase structure that may be emotional one by dependency parsing.Analyzing the word or phrase structure,matching it with the commendatory terms dictionary,derogatory term dictionary,negative word dictionary and adverb of degree dictionary which are included in emotional dictionary,then doing numeration of'sentence weighting to get sentence's emotional polarity.Finally,the sentiment tendency of the whole comment document will be judged by analyzing the polarity of each sentence.(3)The construction of a small e-commerce platform based on the SpringMVC framework by combining with the previous two research works.It supports users to add goods to shopping cart and buy goods.The merchants can store goods in the backstage supporter,manage shops and so on.Meanwhile,set a statistical module which can analyze the evaluation of consumer goods in the backstage supporter,so that the merchant can make different decisions for users according to the their preferences.