Development of Intelligent Web Content Summarization for Accessibility

The increasing volume and complexity of online information have posed significant challenges to digital accessibility, especially for individuals with cognitive, visual, or physical impairments who struggle to navigate and comprehend extensive web content. While existing summarization tools offer basic content reduction capabilities, they often fail to meet established accessibility standards or accommodate users relying on assistive technologies. This project seeks to address these shortcomings through the development of an intelligent web content summarization system that integrates accessibility as a core design principle. The system employs a hybrid summarization approach, combining extractive techniques such as TextRank with advanced abstractive methods based on transformer models like BERT and GPT. A structured developmental methodology was adopted, including requirement analysis, iterative system design, implementation, testing, and refinement. Key modules of the system include a web content crawler, text preprocessing engine, natural language processing layer, summarization core, and an accessibility adaptation module, all designed to ensure compatibility with screen readers and compliance with the Web Content Accessibility Guidelines (WCAG). The system also allows users to customize summary length and language complexity to suit individual needs. Evaluation was conducted using performance efficiency, semantic accuracy, readability, and accessibility-focused metrics. The project ultimately contributes to the broader goal of digital inclusion by enabling more equitable access to information through AI-enhanced, user-centered summarization.