Revisiting General Map Search via Generative Point-of-Interest Retrieval
Dong Chen, Shuai Zheng, Haoyang Shao, Hongsheng Wu, Muhao Xu, Yeyu Yan, Ruifang Li, and Zhenfeng Zhu · 2026
Read Revisiting General Map Search via Generative Point-of-Interest Retrieval by Dong Chen, Shuai Zheng, Haoyang Shao, Hongsheng Wu, Muhao Xu, Yeyu Yan, Ruifang Li, and Zhenfeng Zhu (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-2605-03397.txt.
Description
Dong Chen, Shuai Zheng, Haoyang Shao, Hongsheng Wu, Muhao Xu, Yeyu Yan, Ruifang Li, and Zhenfeng Zhu recast map search as generative retrieval of places. Vague, context-heavy queries defeat classic matching against huge place databases. Intent must be inferred from mixed personal and spatial context.
Their GenPOI framework encodes each place as a token sequence. The sequence carries both meaning and geography. A large language model then generates candidates, under proximity constraints that keep results valid and nearby. Tests used Tencent Map data with over ten million places. The generative approach beat conventional retrieval on hard general search queries.
Catalog record
| CREATOR | Dong Chen, Shuai Zheng, Haoyang Shao, Hongsheng Wu, Muhao Xu, Yeyu Yan, Ruifang Li, and Zhenfeng Zhu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03397