wsdm-reproducibility
Use when hardening the reproducibility of a WSDM paper built on logs, graphs, or user-interaction data - provenance of behavioral datasets, temporal split discipline, click-bias assumptions, seed and variance reporting, privacy-preserving release, and honesty tiers for results no outsider can rerun.
Profile →npx skills add brycewang-stanford/Awesome-Journal-Skills --skill wsdm-reproducibility --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# WSDM Reproducibility Make a web-search/data-mining paper re-derivable. WSDM has no reproducibility checklist to fill (none surfaced for current editions; 待核实 each cycle) - which raises the bar rather than lowering it, because reviewers apply the norm without giving you a form to hide behind. The venue-specific twist: WSDM evidence usually comes from *behavioral* data (queries, clicks, follows, purchases), and behavioral data has failure modes that generic ML reproducibility advice never mentions. ## Provenance: behavioral data decays A log dataset is a measurement of a platform at a moment - the platform's ranker, UI, and user base are all baked into it. Reproducibility therefore starts with recording what generated the data: - Collection window, platform surface (web vs app, market/locale), and any known ranker or UI changes inside the window. - The logging policy: what produced the exposures users could click on. A click log is a *logged-policy* artifact; results on it are conditional on that policy (this is the entire lesson of the position-bias and unbiased learning-to-rank literature born at this venue). - Filtering steps with counts at every stage: bots removed, sessions se
- Provenance: behavioral data decays
- Temporal discipline
- Bias assumptions are part of the method
- Runs, seeds, and variance
- The honesty ladder for unrerunnable results
- Privacy is a reproducibility constraint, not an excuse
- Pre-submission reproducibility sweep
- Output format
What does the wsdm-reproducibility skill do?
Use when hardening the reproducibility of a WSDM paper built on logs, graphs, or user-interaction data - provenance of behavioral datasets, temporal split discipline, click-bias assumptions, seed and variance reporting, privacy-preserving release, and honesty tiers for results no outsider can rerun.
How do I install it?
Run `npx skills add brycewang-stanford/Awesome-Journal-Skills --skill wsdm-reproducibility --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Awesome-Journal-Skills, a repository with 909 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.