Where on Earth Do Users Say They Are?: Geo-Entity Linking for Noisy Multilingual User Input
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914774957162496 |
|---|---|
| author | Masis, Tessa O'Connor, Brendan |
| author_facet | Masis, Tessa O'Connor, Brendan |
| contents | Geo-entity linking is the task of linking a location mention to the real-world geographic location. In this paper we explore the challenging task of geo-entity linking for noisy, multilingual social media data. There are few open-source multilingual geo-entity linking tools available and existing ones are often rule-based, which break easily in social media settings, or LLM-based, which are too expensive for large-scale datasets. We present a method which represents real-world locations as averaged embeddings from labeled user-input location names and allows for selective prediction via an interpretable confidence score. We show that our approach improves geo-entity linking on a global and multilingual social media dataset, and discuss progress and problems with evaluating at different geographic granularities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_18784 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Where on Earth Do Users Say They Are?: Geo-Entity Linking for Noisy Multilingual User Input Masis, Tessa O'Connor, Brendan Computation and Language Artificial Intelligence Geo-entity linking is the task of linking a location mention to the real-world geographic location. In this paper we explore the challenging task of geo-entity linking for noisy, multilingual social media data. There are few open-source multilingual geo-entity linking tools available and existing ones are often rule-based, which break easily in social media settings, or LLM-based, which are too expensive for large-scale datasets. We present a method which represents real-world locations as averaged embeddings from labeled user-input location names and allows for selective prediction via an interpretable confidence score. We show that our approach improves geo-entity linking on a global and multilingual social media dataset, and discuss progress and problems with evaluating at different geographic granularities. |
| title | Where on Earth Do Users Say They Are?: Geo-Entity Linking for Noisy Multilingual User Input |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2404.18784 |