2026.semeval problems

We are seeing many corrections and complaints about 2026.semeval (#9167 #9020 #9118 #9023 #9190), pointing to deeper problems.

I asked copilot to look over the volume and create a checklist. Here is what it gave me below, with some manual edits.

# Audit and correction checklist for `2026.semeval-1`

A comparison of all 455 paper PDFs against their Anthology metadata found several source-data and PDF-association errors.

## Recommendation

- **Do not re-ingest the volume.**
- Process the findings below as coordinated in-place corrections, preserving all existing Anthology IDs and DOIs.
- Ask the SemEval organizers for the unavailable camera-ready PDFs and clarification about two omitted accepted papers.

## Swapped PDFs: `.59` and `.60`

The PDFs for these two papers were swapped in the proceedings source.

- [ ] Replace `2026.semeval-1.59` with the HCMUS RepeatedGames paper currently associated with `.60`.
- [ ] Replace `2026.semeval-1.60` with the UIT-AMMC paper attached to #9118.
- [ ] (optional) Regenerate both PDFs with their corrected Anthology page numbers.
- [ ] Close #9118 after correcting both papers, not only `.60`.

## Wrong or duplicate PDFs

- [ ] `2026.semeval-1.95`: replace the PDF, which currently duplicates `.341`, with the HCMUSDroneBoys camera-ready paper.
- [ ] `2026.semeval-1.142`: replace the PDF, which currently duplicates the Task 9 overview at `.453`, with the king001 camera-ready paper.
- [ ] `2026.semeval-1.259`: replace the anonymized PDF containing “Anonymous ACL submission” with its camera-ready version.
- [ ] `2026.semeval-1.278`: replace the unrelated SemEval-2025 PDF with the corrected SemEval-2026 source PDF now present as `inputs/papers/312.pdf`.
- [x] `2026.semeval-1.416`: replace the anonymized PDF with the camera-ready attachment from #9167.

## Metadata copied from omitted papers

- [ ] `2026.semeval-1.249`: restore authors **Younghee Jeong, Yixin Zhao, and Hoang Bao Trung Le**.
- [ ] `2026.semeval-1.249`: replace its abstract with the abstract from its PDF.
- [ ] `2026.semeval-1.434`: restore author **Vivek Dhayal**.
- [ ] `2026.semeval-1.434`: replace its abstract with the abstract from its PDF.

These errors arose because metadata from omitted source records `280` and `483` was copied onto the following records, `281` and `484`.

## Omitted accepted papers

Both source records were marked “Accept to main conference” but disappeared from the final `papers.yml`.

- [ ] Obtain and publish paper `280`, **“Ping An at SemEval-2026 Task 9: Data Augmentation, Multitask Fine-tuning, and Model Fusion for Multilingual Polarization,”**
- [ ] Same for `483`, **“SemEval-2026 Task 10: Conspiracy Detection,”**

## Author metadata

- [ ] `2026.semeval-1.9`: change **Juan Villate Lemus** to the PDF form **Juan David Villate**.
- [ ] `2026.semeval-1.110`: add omitted PDF author **Michael Olaolu Arowolo**.

## Title metadata

Update these metadata titles to match the camera-ready PDFs:

- [ ] `2026.semeval-1.39`: **CodeHunters at SemEval-2026 Task 13: Detecting Machine-Generated Code by Using Different Transformer Models**
- [ ] `2026.semeval-1.110`: **BAHAHA at SemEval-2026 Task 1: Multi-Style Humor Generation via Comedian-Inspired Prompting and LLM Self-Assessment**
- [ ] `2026.semeval-1.115`: **Team JCT 2026 at SemEval-2026 Task 5: AmbiStory: Navigating Narrative Consistency through a Hybrid LLM-NLI Ensemble**
- [ ] `2026.semeval-1.145`: **IITKanBDone at SemEval-2026 Task 8: ELSER-Based Sparse Retrieval with Mistral-7B for Multi-Turn RAG Evaluation**
- [ ] `2026.semeval-1.175`: update to the full PDF title beginning **“AILS-NTUA at SemEval-2026 Task 8: Query Diversity via Nested Reciprocal Rank Fusion…”**
- [ ] `2026.semeval-1.218`: update to the PDF title beginning **“HausaNLP at SemEval-2026 Task 7: A Locale-Conditional, Multi-Model Prompting Pipeline…”**
- [ ] `2026.semeval-1.333`: **UPR at SemEval-2026 Task 9: Multi-Label Polarization and Manifestation Detection in Urdu**
- [ ] `2026.semeval-1.336`: **MoodMetric at SemEval-2026 Task 4: Dense Transformer Networks for Narrative Story Similarity and Representation**
- [ ] `2026.semeval-1.342`: **GUNLP at SemEval-2026 Task 10: Emotion-Aware Multi-Task Learning for Conspiracy Detection**
- [ ] `2026.semeval-1.388`: **Team p1 at SemEval-2026 Task 5: Semantic Bridge: Augmented Encoding for Word Sense Plausibility**
- [ ] `2026.semeval-1.424`: **Truth Gradient at SemEval-2026 Task 10: Mean Pooling and Narrative Density for Conspiracy Belief Detection**

## Final validation

- [ ] Preserve all existing Anthology IDs and DOIs.
- [ ] Preserve all metadata corrections and revisions made since the original ingestion.
- [ ] Check every replacement PDF’s title and authors against its metadata.
- [ ] Check every replacement PDF’s page count and printed page numbers.
- [ ] Confirm `.95` no longer duplicates `.341`.
- [ ] Confirm `.142` no longer duplicates `.453`.
- [ ] Search all corrected PDFs for “Anonymous ACL submission”.
- [ ] Load the collection through the `acl_anthology` public API.
- [ ] Run the relevant XML/data validation checks.