Datasets:

taqbaylit
/

kabyle-named-entities

language:
  - ber
  - kab
  - en
license: cc-by-4.0
size_categories:
  - n<1K
task_categories:
  - token-classification
  - translation
tags:
  - kabyle
  - berber
  - tamazight
  - named-entity-recognition
  - low-resource-nlp
pretty_name: Kabyle Standardized Named Entities Dataset
dataset_info:
  features:
    - name: kabyle_standardized
      dtype: string
    - name: english_translation
      dtype: string
    - name: entity_category
      dtype: string
  splits:
    - name: train
      num_bytes: 35000
      num_examples: 292
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/kabyle_named_entities.tsv
    csv_kwargs:
      delimiter: "\t"
      quotechar: '"'
      skipinitialspace: true

Kabyle Standardized Named Entities Dataset

This is a manually curated parallel corpus in Kabyle complete with semantic English contextual translations and structured Named Entity Recognition (NER) tag assignments.

Dataset Structure

kabyle_standardized: Target entity string conforming to standardized orthographic regulations.
english_translation: High-context semantic meaning, institutional purpose, or micro-topographic geographical breakdowns.
entity_category: Uniform class label mappings (Location, Political Party, Institution, Movement, Person, Organization).

Source

This work is based on Mohammed Belkacem's initial work on named entities in the Kabyle language.