Label System Of Data Based On Crowdsourcing

As everyone knows, natural language processing and emotion analysis are the very import links in machine study. And these two links need much data as the basic. But most data are acquired from web directly. The quality of the data is irregularity. So we have to label and process these data, then these data can be used in scientific research. But the number of the data obtained from web is very large. It is very onerous to label and process these data only by researchers. We find that as the development of internet, nearly everyone visits the internet every day. And everyone has their major areas. If we can let netizen label and process these data, the work will be easy. This working form is called crowdsourcing. A new concept springs up in the last few years. It will be easier and more convenient for the researchers to get the data of value. This project is based on this requirement. It provides a platform for the users to label and comment. Meanwhile, the platform also recommends users relevant content according to the data they often label and comment. This makes the system more convenient for the user to label and comment.This thesis proposes crowdsourcing as the solution for the system according to the requirements after reading many relevant materials from domestic and oversea. The system uses J2ee as the technology for developing the web system, Mysql as the databse, Tomcat as the server and uses the Bootstrap and Ajax as the technology for the webpage to realize the special effect and visualization. This makes the system more beautiful and easier to use.This system and the relevant documents are achieved independently by analysizing the requirements and system design. All the test results meet the requirements. The system can come into use.