innbyggjarnamn / README.md
titae's picture
Upload dataset
49acf03 verified
metadata
dataset_info:
  - config_name: default
    features:
      - name: Kommune/område
        dtype: string
      - name: Fylke/område
        dtype: string
      - name: Innbyggjarnamn
        list: string
    splits:
      - name: validation
        num_bytes: 17273
        num_examples: 372
      - name: test
        num_bytes: 17273
        num_examples: 372
    download_size: 20604
    dataset_size: 34546
  - config_name: original
    features:
      - name: Innbyggjarnamn
        dtype: string
      - name: Kommune/område
        dtype: string
      - name: Fylke/område
        dtype: string
      - name: Merknad
        dtype: string
    splits:
      - name: test
        num_bytes: 36050
        num_examples: 821
    download_size: 17206
    dataset_size: 36050
configs:
  - config_name: default
    data_files:
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
  - config_name: original
    data_files:
      - split: test
        path: original/test-*

Innbyggjarnamn (Demonyms)

This dataset contains demonyms from different places in Norway. It is based on the Innbyggjarnamn table from the Språkrådet website: https://web.archive.org/web/20250812145750/https://sprakradet.no/stedsnavn-og-navn-pa-statsorgan/navnelister-norsk-skrivemate/innbyggjarnamn/.

License: NLOD 2.0.

The table on the website contains multiple demonyms listed in one row like this "bergensar/-er" or "bam(b)ling", and with multiple rows for some places (with "Merknad" describing the differences between the rows).

In this dataset, all demonyms for a place are listed together, and in full, e.g. ["bergensar", "bergenser"] or ["bamling", "babmbling"].

Load the dataset like this:

from datasets import load_dataset

# Val split
ds = load_dataset("NbAiLab/innbyggjarnamn", split="validation")
# Dataset({
#     features: ['Kommune/område', 'Fylke/område', 'Innbyggjarnamn'],
#     num_rows: 372
# })

# Test split
ds = load_dataset("NbAiLab/innbyggjarnamn", split="test")
# Dataset({
#     features: ['Kommune/område', 'Fylke/område', 'Innbyggjarnamn'],
#     num_rows: 372
# })

To load the original table, load the dataset with name="original":

from datasets import load_dataset

ds = load_dataset("NbAiLab/innbyggjarnamn", name="original", split="test")
# Dataset({
#     features: ['Innbyggjarnamn', 'Kommune/område', 'Fylke/område', 'Merknad'],
#     num_rows: 821
# })
# (some places have multiple rows)