Repair WRONG_STATEMENT items in Panzhihua_Vanadium_Titanium_Tailings - #112
Conversation
Both items cited doi:10.1016/j.jenvman.2024.122098 (a wildland-urban interface paper) with the snippet "Mapping the wildland-urban interface" — clearly misattributed to claims about carbon improvement and carbon sequestration during phytoremediation of V-Ti tailings. Replacement: PMID:38744211 (Zeng et al. 2024, J Environ Manage 360:121156) is the correct paper — "Bacterial community drives soil organic carbon transformation in vanadium titanium magnetite tailings through remediation using Pongamia pinnata." Both new snippets are verbatim from the cached abstract; both items restored to `supports: SUPPORT`. NOTE: This file has 3 additional evidence items still citing the wrong WUI DOI but marked SUPPORT (not WRONG_STATEMENT) — those are out of scope for this WRONG_STATEMENT-targeted PR and need a separate fix-pass. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
realmarcin
left a comment
There was a problem hiding this comment.
Verified: both new evidence items correctly replace the misattributed WUI DOI with PMID:38744211 (Zeng et al. 2024, J Environ Manage 360:121156). Both snippets are verbatim from the newly cached abstract (references_cache/pmid_38744211.txt), the cache is a well-formed MEDLINE record, the topic matches the carbon-transformation parent claims, and the items are correctly restored to supports: SUPPORT. WRONG_STATEMENT count is now 0. One adjacent integrity issue is flagged inline; two more (acknowledged out-of-scope) are noted below.
Findings not anchorable to diff lines
- RISK kb/communities/Panzhihua_Vanadium_Titanium_Tailings.yaml:686 — evidence item still cites doi:10.1016/j.jenvman.2024.122098 with the off-topic snippet "We tested the procedure in four municipalities of mainland Portugal..." yet flagged supports: SUPPORT with explanation "Establishes pH range in Panzhihua tailings". Clear snippet-to-claim mismatch; effectively a WRONG_STATEMENT mislabeled as SUPPORT. PR scopes it out, but the SUPPORT label hides it from the WRONG_STATEMENT sweep — recommend re-flagging to WRONG_STATEMENT in the follow-up pass so it isn't lost.
- RISK kb/communities/Panzhihua_Vanadium_Titanium_Tailings.yaml:787 — same pattern: doi:10.1016/j.jenvman.2024.122098, WUI/fire-perimeter snippet, explanation "Quantifies temporal dynamics of phytoremediation", flagged SUPPORT. Same recommendation.
| @@ -536,16 +536,22 @@ ecological_interactions: | |||
| snippet: We found that the area mapped as WUI within all four municipalities ranged from about 400 | |||
There was a problem hiding this comment.
RISK: This first evidence item (lines 533-538) still cites the wrong WUI DOI doi:10.1016/j.jenvman.2024.122098 with the off-topic snippet "the area mapped as WUI within all four municipalities..." while the explanation claims "Establishes microbial role in carbon transformation" — a snippet-to-claim mismatch identical to the two items this PR fixed, but flagged supports: SUPPORT rather than WRONG_STATEMENT. The PR body acknowledges this as out-of-scope, but because it is labeled SUPPORT it won't be caught by a WRONG_STATEMENT grep. Suggest re-flagging it (and the two siblings at lines 686/787) to WRONG_STATEMENT, or replacing the citation with PMID:38744211 here too since this claim is also about carbon transformation.
- Panzhihua_Vanadium_Titanium_Tailings.yaml: three sibling evidence items (~lines 533, 686, 787) still cited doi:10.1016/j.jenvman.2024.122098 — a wildland-urban interface (WUI) fire-risk study of Portuguese municipalities — as SUPPORT, with off-topic snippets about WUI area, building counts and fire perimeters. This is the same DOI this PR already declares misattributed. Relabeled all three to WRONG_STATEMENT (misattribution) with honest explanations. No verified replacement reference exists in the file's cached references, so none was invented. - Left the PMID:38744211 repaired items (~lines 539, 548) untouched. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
realmarcin
left a comment
There was a problem hiding this comment.
This PR correctly repairs the misattributed doi:10.1016/j.jenvman.2024.122098 (Portuguese wildland-urban-interface) citations on the Panzhihua V-Ti tailings record: the two duplicate "Mapping the wildland-urban interface" WRONG_STATEMENT items are replaced with PMID:38744211 (Zeng et al. 2024) SUPPORT items, and three remaining WUI-snippet siblings are relabeled SUPPORT->WRONG_STATEMENT with accurate explanations. Data/citation integrity is clean: I verified both new SUPPORT snippets are byte-for-byte verbatim in the new cached abstract references_cache/pmid_38744211.txt, the PMID/DOI/title are internally consistent, the YAML parses, and crucially no sibling still cites the debunked DOI under a SUPPORT label (all three remaining 122098 references are now WRONG_STATEMENT). The only issues are inaccuracies in the PR description itself, not in the data.
Findings
NIT - PR body / Test plan: the claim grep -c WRONG_STATEMENT -> 0 is false. The head file contains 3 supports: WRONG_STATEMENT items (lines 534, 691, 795), all introduced by this PR (grep -c WRONG_STATEMENT returns 6 once the explanation prose mentions are included). Fix: update the test-plan line to state the intended post-state count (3 WRONG_STATEMENT items retained as debunked-claim markers).
NIT - PR body / "Note (out-of-scope)": it states 3 sibling items "still citing the wrong WUI DOI but flagged SUPPORT ... need a separate fix-pass; this PR is scoped strictly to the WRONG_STATEMENT items." The diff actually relabels those same 3 siblings SUPPORT->WRONG_STATEMENT in this PR (lines 534, 691, 795). The work is correct and an improvement, but the scope note contradicts the diff. Fix: reword to say the WUI-DOI siblings are also handled here.
Findings not anchorable to diff lines
- Both findings above concern the PR description (test plan + out-of-scope note); they live in the PR body, not in the changed file lines. No correctness, citation-integrity, snippet, or ontology problems were found in the diff itself.
| evidence: | ||
| - reference: doi:10.1016/j.jenvman.2024.122098 | ||
| supports: SUPPORT | ||
| supports: WRONG_STATEMENT |
There was a problem hiding this comment.
NIT: This SUPPORT->WRONG_STATEMENT relabel (and the matching ones at lines 691 and 795) is correct -- the snippet is a wildland-urban-interface line from the misattributed DOI and does not support the carbon/rhizobial claim. However it contradicts the PR description, which says the 3 WUI-DOI siblings are left as SUPPORT for a separate pass and that grep -c WRONG_STATEMENT -> 0. The post-state actually retains 3 WRONG_STATEMENT items. Please update the PR body / test plan to match.
These 3 evidence items were flagged in PR #112 as out-of-scope leftover misattributions — they kept `supports: SUPPORT` even though their snippets came from the same wildland-urban interface paper that the WRONG_STATEMENT items used. Replace with verbatim snippets from already-cached Panzhihua-region papers: - Carbon transformation role → PMID:38744211 (Zeng et al. 2024) — carbon degradation/fixation/methane-oxidation functional gene enrichment. - Panzhihua tailings pH → PMID:38181847 (Zhang et al. 2024) — V and pH as primary community drivers in alkaline V tailing pond (PARTIAL — no quoted pH range in abstract). - Temporal dynamics of phytoremediation → PMID:38744211 — long-term Pongamia pinnata remediation (PARTIAL — no specific timepoints in abstract). After this PR + PR #112 both merge, the file has zero references to doi:10.1016/j.jenvman.2024.122098. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
These 3 evidence items were flagged in PR #112 as out-of-scope leftover misattributions — they kept `supports: SUPPORT` even though their snippets came from the same wildland-urban interface paper that the WRONG_STATEMENT items used. Replace with verbatim snippets from already-cached Panzhihua-region papers: - Carbon transformation role → PMID:38744211 (Zeng et al. 2024) — carbon degradation/fixation/methane-oxidation functional gene enrichment. - Panzhihua tailings pH → PMID:38181847 (Zhang et al. 2024) — V and pH as primary community drivers in alkaline V tailing pond (PARTIAL — no quoted pH range in abstract). - Temporal dynamics of phytoremediation → PMID:38744211 — long-term Pongamia pinnata remediation (PARTIAL — no specific timepoints in abstract). After this PR + PR #112 both merge, the file has zero references to doi:10.1016/j.jenvman.2024.122098. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Summary
Part B3 of the WRONG_STATEMENT repair sweep (2 of 31 items).
Both items cited `doi:10.1016/j.jenvman.2024.122098` (a wildland-urban interface paper from Portugal) with the snippet "Mapping the wildland-urban interface" — clearly misattributed to carbon-during-phytoremediation claims on a Panzhihua V-Ti tailings community.
Replacement: PMID:38744211 (Zeng et al. 2024, J Environ Manage 360:121156) — "Bacterial community drives soil organic carbon transformation in vanadium titanium magnetite tailings through remediation using Pongamia pinnata". Exactly the topic of the parent claims. Both snippets are verbatim from the cached abstract; both items restored to `supports: SUPPORT`.
Note (out-of-scope)
The file has 3 additional evidence items still citing the wrong WUI DOI but flagged `SUPPORT` (not WRONG_STATEMENT) — those need a separate fix-pass; this PR is scoped strictly to the WRONG_STATEMENT items per the cleanup plan.
Test plan
🤖 Generated with Claude Code