REANIMATOR addresses a critical challenge in information retrieval evaluation by repurposing existing test collections through enhanced data extraction and synthesis.

• Extracts valuable information from PDF files including full texts, machine-readable tables, and contextual information
• Enhances generalizability across different retrieval tasks by enriching collections with related content
• Demonstrated effectiveness using the TREC-COVID test collection, showing particular value for medical information retrieval
• Enables reuse of existing resources rather than creating entirely new collections

For medical applications, REANIMATOR offers significant potential to improve retrieval systems for clinical research, enabling faster, more accurate access to relevant medical literature during time-critical situations like pandemics.

REANIMATOR: Reanimate Retrieval Test Collections with Extracted and Synthetic Resources