MultiOCR-QA: Dataset for Evaluating Robustness of LLMs in Question Answering on Multilingual OCR Texts DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; July 13–18, 2025; Padova, IT ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Information s
Bhawna Piryani, Jamshid Mozafari, Abdelrahman Abdallah, Antoine Doucet, Adam Jatowt · 2025
Read MultiOCR-QA: Dataset for Evaluating Robustness of LLMs in Question Answering on Multilingual OCR Texts DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; July 13–18, 2025; Padova, IT ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Information s by Bhawna Piryani, Jamshid Mozafari, Abdelrahman Abdallah, Antoine Doucet, Adam Jatowt (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-16781.txt.
Description
Bhawna Piryani and co-authors introduce MultiOCR-QA in a 2025 paper on question answering over noisy scanned text. Optical character recognition turns historical and multilingual documents into text, but it inserts, deletes, and swaps characters. Those errors can badly hurt systems that answer questions from the text.
Their data set holds fifty thousand question and answer pairs in English, French, and German, built from OCR processed historical documents with varied noise. Tests of leading language models across three error types showed performance drops sharply on noisy text. The authors argue for more noise resilient systems for digitization work.
Catalog record
| CREATOR | Bhawna Piryani, Jamshid Mozafari, Abdelrahman Abdallah, Antoine Doucet, Adam Jatowt |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-16781