Methodology
How this platform works
Transparency about method is the difference between a knowledge platform and a content farm. This page documents every step.
1. Where community experiences come from
Public quit-smoking communities (primarily Baidu's Quit Smoking forum and comparable public platforms) are used as source material only. We read posts to understand real quit experiences — we never mirror, copy or translate posts wholesale.
2. Privacy filtering and de-identification
- Usernames, avatars, UIDs, QQ/WeChat IDs, emails, phone numbers and real names are removed.
- Any other identifiable detail (workplace names, distinctive personal events) is removed or generalized.
- Unrelated, abusive or inappropriate content is discarded.
3. Structuring and deduplication
Each relevant experience is extracted into the same structured schema — smoking history, quit method, withdrawal, triggers, cravings, relapse, outcome, lessons. Duplicates (same post, same story) are detected and merged. Only what the person explicitly stated is recorded; everything else is "Not reported".
4. Editorial synthesis
Published pages are original English editorial content written from the structured records — never verbatim translation of posts. A community report is always attributed as "a community member reported", never as proof.
5. How evidence is evaluated
- Tier 1: WHO, CDC, NIH, NCI, PubMed-indexed systematic reviews, Cochrane, clinical guidelines.
- Tier 2: peer-reviewed journals, professional organizations, major universities.
- Tier 3: high-quality secondary sources.
- Community: experience source only — never scientific evidence.
Every claim carries an evidence-strength label: Strong, Moderate, Limited, Mixed, Insufficient, or Community Report Only. Research records store title, authors, year, journal, DOI/PMID where available, population, method, findings and limitations.
6. Medical safety gate
Content touching medications, pregnancy, mental health, cardiovascular or respiratory disease, cancer or medication interactions enters a medical review queue and is published only with appropriate framing and disclaimers. This platform does not diagnose, prescribe or give individual advice.
7. Quality scoring and updates
Every page is scored on originality, information gain, evidence quality, source quality, experience value, specificity, accuracy, uncertainty handling, internal links, search intent fit, AIO readiness and medical safety. Content below threshold is not published. Research records carry a "last verified" date and are re-checked on a schedule.
8. What we do not do
We do not fabricate experiences, studies, statistics, DOIs, PMIDs or experts. We do not turn one person's story into a population claim, community reports into medical conclusions, correlation into causation, or medication information into prescriptions.