RAG-based Question Answering over Heterogeneous Data and Text

Christmann, Philipp; Weikum, Gerhard

Item

ITEM ACTIONSEXPORT

Add to Basket

Local TagsRelease HistoryDetailsSummary

Released

Paper

RAG-based Question Answering over Heterogeneous Data and Text

MPS-Authors

/persons/resource/persons244397

Christmann, Philipp
Databases and Information Systems, MPI for Informatics, Max Planck Society;

/persons/resource/persons45720

Weikum, Gerhard
Databases and Information Systems, MPI for Informatics, Max Planck Society;

External Resource

No external resources are shared

Fulltext (restricted access)

There are currently no full texts shared for your IP range.

Fulltext (public)

arXiv:2412.07420.pdf
(Preprint), 591KB

Supplementary Material (public)

There is no public supplementary material available

Citation

Christmann, P., & Weikum, G. (2024). RAG-based Question Answering over Heterogeneous Data and Text. Retrieved from https://arxiv.org/abs/2412.07420.

Cite as: https://hdl.handle.net/21.11116/0000-0010-546F-4

Abstract

This article presents the QUASAR system for question answering over
unstructured text, structured tables, and knowledge graphs, with unified
treatment of all sources. The system adopts a RAG-based architecture, with a
pipeline of evidence retrieval followed by answer generation, with the latter
powered by a moderate-sized language model. Additionally and uniquely, QUASAR
has components for question understanding, to derive crisper input for evidence
retrieval, and for re-ranking and filtering the retrieved evidence before
feeding the most informative pieces into the answer generation. Experiments
with three different benchmarks demonstrate the high answering quality of our
approach, being on par with or better than large GPT models, while keeping the
computational cost and energy consumption orders of magnitude lower.