Research And Implementation Of Information Crowdfunding System For A Telecom Operator

With the development of economic globalization,the competition between enterprises is becoming increasingly fierce.How to attract users and ensure the competitiveness of their products in the market has become the focus of enterprise research.A basic idea is to attract users through continuous optimization of products,improve users' experience.In order to optimize the product,the enterprise needs to obtain the experience information of the product from the users,whom are so scattered that it is difficult to collect the information.With the rapid development of the Internet,enterprises can crowd-raise users' information on the use of enterprise products by building Internet tools.The paper describes the research and implementation process of information crowdfunding system.Software development requirements come from a telecom carrier enterprise.After detailed analysis of the needs,the system functional requirements and non-functional requirements have been determined.Then the hierarchical structure design,network structure design,database design,functional module design and detailed design are made to the system.Under the guidance of system design,this paper describes the detail design and implementation.The user's comment information has become an important source for various businesses to optimize products and services.How to classify the comment information emotionally has become a hot spot at present.This article has carried on the research to this question,proposing the emotion analysis method based on word2vec extended sentiment dictionary and SVM.Firstly,reviews were taken as the raw data for data preprocessing,and then based on the constructed basic sentiment dictionary,word2vec was used to extend the sentiment dictionary.Emotion scores of reviews are calculated based on an extended emotion dictionary.In the end,text features such as positive and negative emotional words,negative words,degree adverbs,turning words,emotion scores and so on are incorporated into the emotion dictionary information.After different combinations,these text features are trained and tested using SVM(Support Vector Machine)to find the optimal combinations,which shows that the proposed emotion analysis method based on word2vec extended sentiment dictionary and SVM is effective.