Automating Error Identification and Evaluating Web Accessibility for Differently Abled Users

In today’s world every aspect of life is managed online; let it be travel, medicine, entertainment, professional, or any other basic amenities. Therefore, successfully navigating the concerned website becomes a big task, especially for people with disabilities. Therefore, to try and make their lives a little easier, W3C gives a set of guidelines for the developers and authors of any website or web application to help people with disabilities to navigate any virtual asset with ease. To accommodate every possible disability, the guidelines are divided into 4 principles, 13 guidelines, and around 78 success criteria which are to be tested for every single page of any website that’s being audited. This becomes a very tedious task as most of the websites have quite a lot of web pages each web page has hundreds of lines of source code and is very difficult to audit manually. There are some tools that pinpoint the errors in the source code but the problem remains, the auditor has to go to every error location and note down the specific path to it which is very time-consuming, the time that could be used to actually rectify these errors. In this paper, we aim to provide a solution to this problem by using Natural Language Processing to find the errors and warning signs and record the position of those errors in the module by giving their source code’s page name and line number in the form of a text file.