1 paper · 1 filter
Joongmin Shin, Gyuho Shim, Jung-hun Lee +1
Open-domain question answering (ODQA) over multimodal document corpora requires linking evidence scattered across text, tables, and figures. Existing systems often store these sour…