Using Semi-supervised Group Sparse Regression to Improve Web Accessibility Evaluation
| Autor: | Liangcheng Li, Lianjun Dai, Jiajun Bu, Wei Wang, Yue Wu, Zhang Yueqing, Zhi Yu |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Information retrieval
Computer science Group (mathematics) 05 social sciences Disabled people Feature (computer vision) Global distribution 0502 economics and business Web page 050211 marketing 0501 psychology and cognitive sciences Evaluation result 050107 human factors Sparse regression Web accessibility |
| Zdroj: | Lecture Notes in Computer Science ISBN: 9783319942766 ICCHP (1) |
| DOI: | 10.1007/978-3-319-94277-3_10 |
| Popis: | Web accessibility evaluation checks the accessibility of the website to help improve the user experiences for disabled people. Due to the massive number of web pages in a website, manually reviewing all the pages becomes totally impractical. But the complexities of evaluating some checkpoints require certain human involvements. To address this issue, we develop the semi-supervised group sparse regression algorithm which takes advantages of the high precision of a small amount of manual evaluation results along with the global distribution of all the web pages and efficiently gives out the overall evaluation result of the website. Moreover, the proposed method can tell the importance of each feature in evaluating each checkpoint. The experiments on various websites demonstrate the superiority of our proposed algorithm. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|