#USA #Frederick #AWS #PDFix #ITHAKA

PDFix and ITHAKA: Transforming PDF Accessibility

In recent times, accessibility in digital content has emerged as a vital concern, especially for educational institutions, libraries, and nonprofit organizations. A remarkable case study highlights this challenge and solution involving PDFix and ITHAKA, an organization known for JSTOR, which has taken significant steps to ensure accessible scholarly content.

A Need for Scalable Solutions

Amidst an extensive collection of approximately 20 million PDFs, the demand arises for a more efficient way to make these documents accessible to researchers, students, and educators. Traditional manual remediation methods posed a significant financial burden, with estimates for completely converting ITHAKA's archives reaching between $156 million and $624 million. Thus, the pressure was on for them to find a new approach to accessibility.

The On-Demand Pipeline Solution

Partnering with Amazon Web Services (AWS), ITHAKA embarked on creating a robust pipeline that streamlines PDF accessibility. Rather than attempting to remediate the entire collection upfront, they embraced an innovative on-demand model. This allows documents to be processed as users request accessible versions rather than in bulk, ensuring resources are allocated efficiently based on actual user needs.

How It Works

With the employment of PDFix, a solution that automates structural tagging, the process has become remarkably efficient. PDFix applies essential structural tags—such as headings, paragraphs, lists, and tables—to individual PDF pages. Additionally, Amazon Bedrock plays a crucial role by generating alternative text descriptions for images and charts. The entire workflow is orchestrated using AWS Step Functions, ensuring that every document passes through rigorous checks for PDF/UA accessibility compliance via validation tools like veraPDF.

Positive Outcomes Achieved

Initial results from this collaboration have been impressive:

• - A staggering 98% pass rate on accessibility checks.
• - Processing costs cut down to an estimated $0.026 per page.
• - Overall, more than 97% cost reduction compared to manual methods.
• - Rapid deployment was achieved just two months after initial workshops conducted by AWS and ITHAKA.

David Herr, the President of PDFix-US, reflected on this transformation, stating, “This implementation demonstrates what becomes possible when PDF accessibility is treated as a scalable technology and workflow challenge.”

The Modular Approach

Notably, ITHAKA's implementation is modular. This means they can integrate new technologies much quicker without overhauling existing structures. If future innovations in PDF accessibility or generative AI arise, ITHAKA is poised to incorporate such advancements seamlessly into their existing pipeline. This not only positions them as forward-thinking but also provides a solid framework for ongoing improvement.

A Model for Others

The success of this initiative serves as a valuable model for other universities, libraries, government agencies, nonprofits, and organizations handling extensive collections of inaccessible documents. Instead of facing an unsustainable manual-remediation expense or leaving vast document collections untouched, PDFix provides the tools necessary to construct scalable document-processing workflows while still ensuring crucial human oversight where needed. Herr remarked, “PDFix gives developers and organizations the technology to build accessibility directly into scalable document-processing workflows.”

Through this collaboration, ITHAKA and PDFix are setting lofty standards for digital content accessibility, demonstrating that both innovation and compassion can coexist in technology. The operational transformation not only enhances accessibility but also exemplifies how smart workflow solutions can drive down costs and enhance service delivery.