Skip to content

Latest commit

 

History

History
60 lines (43 loc) · 2.8 KB

README-topres19th.md

File metadata and controls

60 lines (43 loc) · 2.8 KB
Raw

TopRes19th dataset

Dataset profile
Original dataset https://doi.org/10.23636/r7d4-kw08
Document type newspaper (19C)
Languages English
Annotation guidelines see README in original release and below
Annotation tool INCEpTION
Original format and tagging scheme WebAnno 2.3
Annotations NERC, EL (towards Wikipedia)
Version (used in HIPE-2022) v2
Related publication A Dataset for Toponym Resolution in Nineteenth-Century English Newspapers
License CC BY-NC-SA 4.0 Attribution-NonCommercial-ShareAlike

Entity tagset

The topres19th dataset focuses exclusively on toponyms.

Coarse-grained tagset Fine-grained tagset Nesting applies Linking applies Present in HIPE 2022 data
LOC _ no yes yes
BUILDING - no yes yes
STREET - no yes yes
ALIEN - no yes no
OTHER - no yes no
FICTION - no yes no

Note: Due to data sparseness the tags ALIEN, OTHER and FICTION and their corresponding annotations (6 in total for the training/dev data) were removed in HIPE 2022.

TopRes19th toponym classes correspond to (descriptions taken from the README of the original dataset release v2):

  • BUILDING: names of buildings, such as the 'British Museum'.
  • STREET: streets, roads, and other odonyms, such as 'Great Russell St'.
  • LOC: any other real world places regardless of type or scale, such as 'Bloomsbury', 'London' or 'Great Britain'.
  • ALIEN: extraterrestrial locations, such as 'Venus'.
  • FICTION: fictional or mythical places, such as 'Hell'.
  • OTHER: other types of entities with coordinates, such as events, like the 'Battle of Waterloo'.

HIPE-2022 Tasks and Challenges

The topres19th dataset can be used for:

  • Tasks: NERC-Coarse, NEL.
  • Challenges: Multilingual Newspaper Coarse, Global Adaptation Coarse.

Specificities and important information

  • Documents: topres19th documents corresponds to newspaper articles.
  • Metadata: topres19th has rich metadata, reported in the HIPE-2022 converted files.
  • Sentence splitting: performed automatically on OCRed text (performances not perfect).
  • Entity linking and metonymic sense: originally towards Wikipedia, mapped to Wikidata during HIPE-2022 conversion.
  • Known glitches:
    • none at present