Web Accessibility Evaluation Based On Checkpoint Sampling

With the rapid development of Internet, the internet has become main way of access to information and services. However, people with disabilities due to their physical limitations, they can’t visit most of web site. So they unable to get the information they wanted from the internet. Although, many organizations published many web content accessibility guidelines, but unfortunately did not change the present situation that the disabilities difficult to obtain information from the internet. So in order to promote the development of web accessibility, we need to evaluate the website’s accessibility situation effectively and urge them into shape.Web accessibility evaluation needs a lot of manual testing which cost is expensive, and difficult apply to large sites. In order to reduce cost, we sampling web pages from site, so sample can deeply influences website’s accessibility evaluation result. Through the analysis of the several current commonly used web page sampling algorithms, we find that these algorithms can’t choose the representative sample, it can deeply influenced the evaluation results. So the paper proposes three web page sampling approaches based on checkpoint. And we designed some contrast experiments, the results show that heuristic sampling checkpoint based algorithm has the optimal result.When some website accessibility evaluation task is urgent, the algorithms proposed before cannot meet the need of online sampling. So we improve three offline checkpoint-based sampling algorithms, make them suitable for online sampling. After repeated experiments, we find that the number of the samples and the arrival of the new web pages can effects sampling result. And heuristic online sampling algorithm relatively has optimal result.Finally, this paper contrasts checkpoint-based offline sampling algorithm and online sampling algorithm, In most cases, the offline sampling algorithms are superior to online sampling algorithms.