metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 130703
num_examples: 250
download_size: 54948
dataset_size: 130703
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
task_categories:
- question-answering
language:
- ja
Japanese construction themes FAQs scraped from https://www.city.yokohama.lg.jp/business/bunyabetsu/kenchiku/annai/faq/qa.html.
Downloaded using the following code:
import requests
from lxml import html
import pandas as pd
from datasets import Dataset
hrefs = [
"/business/bunyabetsu/kenchiku/annai/faq/ji-annnai.html",
"/business/bunyabetsu/kenchiku/tetsuduki/kakunin/qa-kakunin.html",
"/business/bunyabetsu/kenchiku/tetsuduki/teikihoukoku/seido/01.html",
"/business/bunyabetsu/kenchiku/tetsuduki/teikihoukoku/seido/07.html",
"/business/bunyabetsu/kenchiku/tetsuduki/doro/qa-doro.html",
"/business/bunyabetsu/kenchiku/tetsuduki/doro/qa-doro.html",
"/business/bunyabetsu/kenchiku/bosai/kyoai/jigyou/qanda.html",
"/business/bunyabetsu/kenchiku/tetsuduki/kyoka/43.html",
"/business/bunyabetsu/kenchiku/takuchi/toiawase/keikakuho/tokeihou.html",
"/business/bunyabetsu/kenchiku/takuchi/toiawase/kiseiho/takuzo.html",
"/business/bunyabetsu/kenchiku/takuchi/toiawase/keikakuho/q4-1.html",
"/business/bunyabetsu/kenchiku/kankyo-shoene/casbee/hairyo/qa.html",
"/business/bunyabetsu/kenchiku/tetsuduki/jorei/machizukuri/fukumachiqa.html",
"/business/bunyabetsu/kenchiku/kankyo-shoene/chouki/qa-chouki.html",
"/business/bunyabetsu/kenchiku/kankyo-shoene/huuti/qa-huuchi.html",
"/kurashi/machizukuri-kankyo/kotsu/toshikotsu/chushajo/jorei/qa.html",
]
url_stem = "https://www.city.yokohama.lg.jp"
def get_question_text(url):
# Send a GET request to the webpage
response = requests.get(url)
# Parse the HTML content
tree = html.fromstring(response.content)
question_data = []
# Use XPath to find the desired elements
for qa_element in tree.xpath('//div[@class="contents-area"]/section'):
question_data.append({
"question": qa_element.xpath('.//div[@class="question-text"]/text()')[0],
"answer": "\n".join(qa_element.xpath('.//div[@class="answer-text"]/div/p/text()'))
})
return question_data
qa_list = []
for href in hrefs:
print(href)
qa_list.extend(get_question_text(url_stem + href))
df = pd.DataFrame(qa_list)
df.question = df.question.apply(lambda x: x[len(x.split()[0]):] if " " in x[:7] or " " in x[:7] else x)
df.answer = df.answer.apply(lambda x: x[len(x.split()[0]):] if " " in x[:7] or " " in x[:7] else x)
df.question = df.question.str.strip()
df.answer = df.answer.str.strip()
df.question = df.question.apply(lambda x: x[:-len(x.split("＜")[-1])-1] if "＜" in x else x)
df.answer = df.answer.apply(lambda x: x[:-len(x.split("＜")[-1])-1] if "＜" in x else x)
df.question = df.question.str.strip()
df.answer = df.answer.str.strip()
Dataset.from_pandas(df).push_to_hub("lightblue/architecture_faqs")