Preprint
Aug 2026
EnSI-RAG: Entity-Structure-Indexed Retrieval-Augmented Generation for Long-Document Question Answering
This work proposes EnSI-RAG (Entity-Structure-Indexed Retrieval-Augmented Generation), a framework that constructs a query-independent, entity-centered index that separates evidence localization from answer synthesis while preserving traceable source evidence.
Xuanyu Meng, Jiashuo Sun, Jash Parekh et al.
· 0 citations