CHANGELOG v0 10 6
# Changelog — v0.10.6
## Collector
- Versioned all collection outputs to v0.10.6.
- Eligibility and institutional identity/boundary handling unchanged.
- Raw evidence collection logic remains aggregate/count/group based.
- Full Groupings export retained for native KPI20 SDG Diversity scoring.
- Candidate methodology audit sheet synchronized with v0.10.6 weights and KPI10/KPI21/KPI22 definitions.
- Single OpenAlex key retained; no dual-key speed mode added because it does not accelerate the sequential OpenAlex request path.
## Scorer
- Default input synchronized to `HE_Open_Data_Ranking_2027_Metrics_Master_v0_10_6.xlsx`.
- Pillar weights: 50/18/15/7/10.
- KPI weights total 100% / 10,000 points.
- KPI8 remains 600 points.
- KPI10 uses multilateral intensity among international works.
- KPI21 uses SDG excellence advantage versus institutional Top-10 baseline.
- KPI22 uses SDG OA advantage versus institutional OA baseline.
- Eligibility and institutional-boundary handling unchanged.
- Missing/zero/structural-zero audit policy retained.
- Sparse KPI8 positive-only normalization retained.
- Optional `--sdg-diversity-reference` remains only as compatibility support for older raw workbooks that lack Groupings; fresh v0.10.6 collector exports include Groupings.
## Researcher Intelligence package revision — v1.0
- Added `he_researcher_intelligence_v1_0.py` as a non-scoring dashboard module.
- Added HE Institutional Research Contribution Score (HE-IRCS): 40% affiliated citations, 25% affiliated h-index, 15% affiliated i10-index, 10% affiliated publications, 10% recent institutional citations; all components percentile-normalized.
- Added exact author+institution verification inside the same OpenAlex authorship record.
- Added up to 100 Leading Research Contributors per university.
- Added API pilot mode and local OpenAlex JSONL snapshot production mode.
- API work retrieval uses batched OR-filters for candidate authors, reducing duplicated requests.
- Added XLSX, CSV, JSON and SQLite researcher-intelligence outputs.
- Added Windows launchers for API pilot, snapshot production, status, export and score recalculation.
- Core v0.10.6 ranking methodology, 24 KPIs, five pillar weights and 10,000-point score are unchanged.
## Collector
- Versioned all collection outputs to v0.10.6.
- Eligibility and institutional identity/boundary handling unchanged.
- Raw evidence collection logic remains aggregate/count/group based.
- Full Groupings export retained for native KPI20 SDG Diversity scoring.
- Candidate methodology audit sheet synchronized with v0.10.6 weights and KPI10/KPI21/KPI22 definitions.
- Single OpenAlex key retained; no dual-key speed mode added because it does not accelerate the sequential OpenAlex request path.
## Scorer
- Default input synchronized to `HE_Open_Data_Ranking_2027_Metrics_Master_v0_10_6.xlsx`.
- Pillar weights: 50/18/15/7/10.
- KPI weights total 100% / 10,000 points.
- KPI8 remains 600 points.
- KPI10 uses multilateral intensity among international works.
- KPI21 uses SDG excellence advantage versus institutional Top-10 baseline.
- KPI22 uses SDG OA advantage versus institutional OA baseline.
- Eligibility and institutional-boundary handling unchanged.
- Missing/zero/structural-zero audit policy retained.
- Sparse KPI8 positive-only normalization retained.
- Optional `--sdg-diversity-reference` remains only as compatibility support for older raw workbooks that lack Groupings; fresh v0.10.6 collector exports include Groupings.
## Researcher Intelligence package revision — v1.0
- Added `he_researcher_intelligence_v1_0.py` as a non-scoring dashboard module.
- Added HE Institutional Research Contribution Score (HE-IRCS): 40% affiliated citations, 25% affiliated h-index, 15% affiliated i10-index, 10% affiliated publications, 10% recent institutional citations; all components percentile-normalized.
- Added exact author+institution verification inside the same OpenAlex authorship record.
- Added up to 100 Leading Research Contributors per university.
- Added API pilot mode and local OpenAlex JSONL snapshot production mode.
- API work retrieval uses batched OR-filters for candidate authors, reducing duplicated requests.
- Added XLSX, CSV, JSON and SQLite researcher-intelligence outputs.
- Added Windows launchers for API pilot, snapshot production, status, export and score recalculation.
- Core v0.10.6 ranking methodology, 24 KPIs, five pillar weights and 10,000-point score are unchanged.