NamuWikiText
NamuWikiText is a dataset released by lovit@github. It provides Namu Wikipedia in a text format. Data specification is as follows.
- author: lovit@github
- repository: https://github.com/lovit/namuwikitext
- size:
- train: 31,235,096 lines (500,104 docs, 4.6G)
- dev: 153,605 lines (2,525 docs, 23M)
- test: 160,233 lines (2,527 docs, 24M)
Data structure is as follows:
| Attributes | Property |
|---|---|
| text | a body of a section |
| pair | a title of a section |
1. Using in Python
You can download and load the corpus after executing your Python console.
Downloading the corpus
You can download NamuWikiText corpus into your local directory with the following Python codes.
from Korpora import Korpora
Korpora.fetch("namuwikitext")
By default, the corpus is downloaded to a Korpora directory within the user’s root directory (~/Korpora). If you wish to download the corpus to another directory,add root_dir=custom_path argument to the fetch method.
When the fetch method is executed with force_download=True argument, it ignores the existing corpus in the local directory and re-downloads the corpus. The default value of force_download is False.
Loading the corpus
You can load NamuWikiText corpus from your Python console with the following codes. If the corpus does not exist in the local directory, it is also downloaded as well.
from Korpora import Korpora
corpus = Korpora.load("namuwikitext")
You can also load the corpus as follows. The output of these codes is identical to that of previous codes.
from Korpora import NamuwikiTextKorpus
corpus = NamuwikiTextKorpus()
If you use either one of these previous examples, you can load the corpus into the variable corpus. train refers to the training dataset of the corpus, and you can check its first training instance as follows.
>>> corpus.train[0]
SentencePair(text='상위 문서: 아스날 FC\n2009-10 시즌 2011-12 시즌\n2010 -11 시즌...', pair=' = 아스날 FC/2010-11 시즌 =')
>>> corpus.train[0].text
상위 문서: 아스날 FC\n2009-10 시즌 2011-12 시즌\n2010 -11 시즌...
>>> corpus.train[0].pair
= 아스날 FC/2010-11 시즌 =
dev and test refer to the validation and test datasets of the corpus, respectively. Each of their first instance can be accessed as follows.
>>> corpus.dev[0]
SentencePair(text='상위 항목: 축구 관련 인물, 외국인 선수/역대 프로축구\n...', pair=' = 소말리아(축구선수) =')
>>> corpus.test[0]
SentencePair(text='', pair=' = 덴덴타운 =')
By executing the get_all_texts method, you can access all texts (bodies of sections) within the corpus.
>>> corpus.get_all_texts()
['상위 문서: 아스날 FC\n2009-10 시즌 2011-12 시즌\n2010 -11 시즌...', ... ]
By executing the get_all_pairs method, you can access all pairs (titles of sections) within the corpus.
>>> corpus.get_all_pairs()
['= 아스날 FC/2010-11 시즌 =', ... ]
2. Using in a terminal
You can directly download the corpus without executing Python console. To do so, use the following command.
korpora fetch --corpus namuwikitext
By default, the corpus is downloaded to a Korpora directory within the user’s root directory (~/Korpora). If you wish to download the corpus to another directory,add --root_dir custom_path argument to the fetch command.
If you add --force_download argument when executing the fetch command in the terminal, it ignores the existing corpus in the local directory and re-downloads the corpus.