Yahoo! Chiebukuro Q&A Scraper

Collect public Yahoo!知恵袋 questions, answers, best-answer decisions, replies, author labels, timestamps, engagement signals, categories, and source URLs for Voice of Customer (VOC), market research, and Japanese-language community analysis.

Use this Actor when the research unit is a complete question context: the question, its selected best answer, and the other public answers.

How it relates to Quora scrapers

The input and output design follows the useful parts of established Quora Actors: keyword search and direct URLs can be mixed, public Q&A content is structured for downstream analysis, and question/answer records retain source metadata.

Yahoo!知恵袋 is not a threaded forum like Reddit or 贴吧. It is a question-first service closer to Quora or 百度知道. This Actor therefore preserves the relationship between each question and all of its answers instead of treating answer text as context-free comments.

Capability	Typical Quora scraper	This Actor
Keyword search	Yes	Yes, using public Yahoo!知恵袋 search pages
Direct question URLs	Yes	Yes
Questions and answers	Yes	Yes
Best/accepted answer	Accepted answer when exposed	Yahoo!知恵袋 ベストアンサー flag and questioner thanks
Reply context	Varies	Public replies nested under their answer when exposed
Flat CSV-friendly rows	Often	Yes; each answer repeats its question context
Coverage audit	Varies	Search-page, question-level, and run-level audit records
Profiles, topics, spaces, posts	Common on Quora	Not applicable; this Actor is deliberately limited to Yahoo!知恵袋 Q&A

When to use it

Choose this Actor when:

• You need Japanese consumer questions, purchase barriers, usage problems, recommendations, or peer-proposed solutions.
• You want to search several Japanese keywords and deduplicate overlapping questions.
• You already have Yahoo!知恵袋 question URLs and need the complete public Q&A context.
• You need best-answer status, answer reactions, public author labels, categories, and reply context.
• You need audit rows that show search coverage, answer coverage, failures, and stopping conditions.

Choose another workflow when:

• You need Reddit-style comment trees or community/subreddit discovery.
• You need private, deleted, authenticated, account-specific, or historical content that is no longer publicly rendered.
• You need Yahoo! JAPAN Search results outside Yahoo!知恵袋.
• You need semantic analysis or AI summaries instead of source records.

Input

Use searchQueries, startUrls, or both. Questions found by several searches are collected once and retain every matching query/rank in search_hits.

{
  "searchQueries": [
    "防災グッズ おすすめ",
    "電気代 節約 方法",
    "キャンプ 初心者"
  ],
  "startUrls": [
    {
      "url": "https://detail.chiebukuro.yahoo.co.jp/qa/question_detail/q10329103845"
    }
  ],
  "questionStatus": "all",
  "maxQuestionsPerQuery": 100,
  "maxAnswersPerQuestion": 0,
  "includeReplies": true,
  "outputMode": "both",
  "concurrency": 5,
  "requestDelayMs": 250
}

Main controls

Field	Meaning
searchQueries	Japanese keyword searches.
startUrls	Direct public question detail URLs.
questionStatus	all, open, or resolved; applies to searches only.
maxQuestionsPerQuery	Per-query result cap before cross-query deduplication.
maxSearchPages	Hard safety cap; Yahoo currently renders up to 10 search results per page.
maxAnswersPerQuestion	0 means every displayed answer; a positive value creates a bounded sample.
includeReplies	Keeps replies nested under answers when Yahoo exposes them in the public page data.
outputMode	nested, flat, or both. Default both supports JSON context and CSV/Excel analysis.
concurrency	Parallel question requests, 1-10.
requestDelayMs	Delay before each Yahoo request.
maxRequestRetries	Total attempts for transient failures.

Output

The default dataset contains five record types. Always filter record_type before analysis.

record_type	One row per	Purpose
question	Unique question	Full question metadata and, in nested/both mode, all collected answers.
answer	Answer	CSV-friendly answer row with repeated question title, text, categories, URL, and search provenance.
search_audit	Query page	Query, page, results found, new unique questions, displayed total, and errors.
question_audit	Question	Displayed answer target, collected answers/pages/replies, coverage flags, and errors.
run_audit	Actor run	Input counts, unique questions, output totals, failures, and completion time.

For VOC analysis, use either:

• question rows in nested mode, where each question carries its answers; or
• answer rows in flat/both mode, where every answer repeats question_id, question_title, question_text, and source context.

Stable deduplication keys are question_id and answer_id.

Pricing

The question-completed event is charged once after a question has been parsed successfully, its requested answer target has been reached, and its question and audit rows have been written to the Dataset. Failed questions, incomplete answer coverage, answer rows, replies, and audit rows are not charged separately.

The run_audit row reports questions_chargeable, question_completed_events_charged, and any questions_not_charged because the run reached the user's maximum charge limit.

Coverage checks

Before analysis, verify:

1. run_audit.failures is empty.
2. Every question_audit.answer_coverage_complete value is true, unless a deliberate answer cap was used.
3. reply_coverage_complete is true when replies are required. A false value means Yahoo displayed more replies than were present in the server-rendered page data.
4. Search-page counts and per-query limits match the intended research scope.

Limits and responsible use

• Collects public Yahoo!知恵袋 pages only and does not log in or bypass access controls.
• Page structure, field availability, and anti-bot behavior can change; audit failures must be reviewed.
• Public usernames can be masked by Yahoo. The Actor preserves only the public display value and does not infer identity.
• Direct answer anchors are constructed from public question and answer IDs; Yahoo may change anchor behavior.
• The Actor does not classify relevance, commercial content, sentiment, or VOC topics.
• Use reasonable request rates and follow applicable law, Yahoo! JAPAN terms, and your organization's data-handling rules.