A University of Washington-led research team has developed GovScape, a new AI-powered search platform that helps users quickly find information within millions of government documents preserved in the End of Term Web Archive.
The tool supports keyword, semantic and visual searches, allowing users to locate documents by specific terms, topics or features such as redactions, charts and photographs. Currently indexing approximately 10 million federal government PDFs, GovScape is designed to make vast digital archives more accessible to researchers, journalists and the public.
The combining of text and image analysis with highly efficient AI models allows the platform can process and search large collections of complex documents at a fraction of the cost of traditional approaches. Researchers plan to expand GovScape to cover the archive’s full collection of 70 million documents, further improving public access to government information and historical records.