Analyzing And Designing Of User Tagging System Of Crowdtest Platform

As improving their machine learning algorithms has become more and more important and tagged data is the basis of doing that, the teams of the company have rapidly increasing needs for publishing all kinds of tagged tasks. Before User Tagging System, most tasks are done by product managers and outsourcing personnel. Because of the limitation of manpower, PMs cannot produce enough tagged data. Besides, the tasks aren’t their decent work too. As to outsourcing personnel, it costs too much to train them and pay. User Tagging project is established to solve these problems mentioned above. Its goal is to make full use of the labor force of the netizens to provide a low-costing and high-profit human resource solution to gain a great number of tagged data for the teams, thus helping them to improve their products’quality.User Tagging System is a sub-system of Crowdtest Platform. It’s developed with MVC designing pattern which is very popular among Web applications. Based on LAMP structure, it’s deployed on Baidu Application Engine. And some other basic functions provided by Baidu Cloud Environment are also used such as cloud storage and cloud database. Developers can be more focused on their business logic and they don’t have to deploy their physical environment and maintain their web server machines. The processing and storage ability can be expanded by simply applying for more processing or storage nodes. But there are some disadvantages too, such as being dependent on the stability of the cloud environment and not able to change the deployment of physical machines according to their own needs.The thesis starts with introducing the background and the application prospect of user tagging system. Then the technology selection is described. The emphasis of the thesis is on the requirement analysis, the designing and the implementation of the system. Concrete functions and some considerations about the designing and implementations are discussed in this part. In the end, the current situation and the future jobs of the system is summarized and forecasted.