Improving Semantic Proximity in Information Retrieval through Cross-Lingual Alignment
Seongtae Hong, Youngjoon Jang, Jungseob Lee, Hyeonseok Moon, Heuiseok LimDepartment of Computer Science and Engineering, Korea University{ghdchlwls123, dew1701, omanma1928, glee889, limhseok}@korea.ac.kr · 2026
Read Improving Semantic Proximity in Information Retrieval through Cross-Lingual Alignment by Seongtae Hong, Youngjoon Jang, Jungseob Lee, Hyeonseok Moon, Heuiseok LimDepartment of Computer Science and Engineering, Korea University{ghdchlwls123, dew1701, omanma1928, glee889, limhseok}@korea.ac.kr (2026) 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-2604-05684.txt.
Description
Hong and four colleagues expose a flaw in multilingual retrieval: when English documents share a pool with another language, retrievers favor unrelated English texts over same-language matches. They build scenarios and metrics for cross-lingual alignment and train a fix on just 2.8k samples. The method lifts cross-lingual scores across most embedding models while curbing the English bias.
Catalog record
| CREATOR | Seongtae Hong, Youngjoon Jang, Jungseob Lee, Hyeonseok Moon, Heuiseok LimDepartment of Computer Science and Engineering, Korea University{ghdchlwls123, dew1701, omanma1928, glee889, limhseok}@korea.ac.kr |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52987 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.05684