The Recommendation System Of Sample Pages For WEB Accessiblity Evaluation

With the rapid development of information technology,people's daily life has been inseparable from the Internet.Internet brings convenience,at the same time,because most sites do not conform to the accessibility guidelines,the disabled can not access to the information conveniently like ordinary people.In order to make the disabled able to browse the site without obstacle,constructing site according to the accessibility guidelines is of great significance.In order to evaluate websites' accessibility and find inaccessibility web pages effectively,web accessibility evaluation is essential.Howerver,the web accessibility evaluation is not fully automated,some rules need manual assessments.Because the web pages is varied and assesments need subjective judgment,the accuracy and efficiency can not be guaranted.In this paper we propose a recommendation method of sample pages for web accessibility evaluation,which can recommendate the sample pages with the highest similarity to evaluators.The sample pages can help usesr understand the rules easier and evaluate the webpages more effectively.We storage the webpage stucture in DOM tree and adopt the improved simple tree matching(STM)to measure the similarity between web documents.We compare the URL to detect the obvious template,which can speed up the calculation of similarity.Furthermore,the similarity of evaluating contents is added to improve the accuracy of recommendation.Finally,the similarity of web page is combinded with structural similarity and the similarity of evaluating contents.Based on the web page similarity algorithm,we propose a method for creating the sample pool for the recommendation system.With the increasment of the sample pages,the new update method is proposed for the high accuracy of recommendation.Finally,the experimental results on real data sets verify the reliability of web page similarity algorithm.And the feedback of users proves that the sample page recommendation algorithm and the update algorithm for sample pool are effective.