metadata
dataset_info:
features:
- name: _id
dtype: string
- name: character_id
dtype: string
- name: article_en
dtype: string
- name: article_vi
dtype: string
- name: character_url
dtype: string
- name: code
dtype: string
- name: description
dtype: string
- name: language
dtype: string
- name: title
dtype: string
- name: type
dtype: string
- name: metadata
dtype: string
- name: summary
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 5843304840
num_examples: 515669
download_size: 3190490993
dataset_size: 5843304840
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
This dataset is used by me using wiki data to get all characters in English and Vietnamese pages in the following list of categories:
- Q5 : human
- Q95074 : fictional_character
- Q15632617 : fictional_human
- Q15773347 : television_character
- Q1114461 : comic_book_character
- Q4271324 : literary_character
Then I use wikipedia-api to get the content of each page characters