Design And Implementation Of A Web Accessibility Evaluation System Based On CrowdSourcing

With the development of information technology and the internet’s being more and more popular, the Internet has been playing an increasingly important role in people’s daily life, study and work. For the ordinary people, the Internet is a very important and convenient way to access and utilize information and service. At the same time, the convenience for the disabled to access the Internet is not paid much attention to. Comparing the population sampling survey in 2006 with it last time, the number and the proportion of the disabled has increased, and the structure of the category for the disabled has changed.China Disabled Persons’ Federation tries to create an environment to help the disabled integrate into the society. They have united with Ministry of industry and information technology, and released the criteria for the information accessibility based on the international accessibility standards in 2012. The criteria stipulates the constraints for web accessibility in perception, understanding, operation and compatibility.The paper addresses the multi-user decision-making solution for artificial detection part and describes the design and implementation of it based on the tasks of web accessibility evaluation system. To check the tasks based on multi-user decision-making solution, we need to solute the following problems:how to allocate tasks, how to handle the tasks for volunteers, and how to judge the final result for multi-user results, in which the most important part for the judging of the final result is multi-user decision-making stage. This paper focuses on it and addresses the solution based on the existing research and practical requirements. Considering the development of crowd sourcing pattern and research of multi-user decision-making problems, This paper adopts the iterative weighted majority voting method based on feedback to rule for the final result among all the volunteers. The method takes users’ historical experience into consideration, with which we can make effective judgment for users’ checking results.