Resource definition
Structured Document Extraction fixture corpus — round 1
A fixed round-1 fixture corpus of 19 fictional PDFs for Structured Document Extraction (R5 v1).
Resource typeDocument set
What this resource provides or represents
This resource provides the pinned document corpus used for Structured Document Extraction evaluations. It represents a fixed set of fictional PDF source documents, so the benchmark can assess extraction from the same public corpus each time.
Scope and composition
The boundaries of the resource and how its meaningful parts are organized for evaluation.
Fixed round-1 corpus for the Structured Document Extraction benchmark.
19 distinct fictional PDFs.
A mix of invoice, receipt and bank-statement documents.
Versions
Public context for the resource versions used in evaluation.
Version1
Version 1
Initial version with 19 PDFs.
Benchmarks that use this resource
A resource has global identity and may be reused across benchmarks.
Structured Document Extraction
Uses Version 1.
→