Add v5 datasheets and semantic evaluations to documentation

- Add new 'Latest: v5 Datasheets and Semantic Evaluations' section
- Include links to all 8 v5 HTML files (4 datasheets + 4 evaluations)
- Document rubric20-semantic evaluation framework and results
- Show evaluation scores: VOICE (96.4%), AI-READI (94.0%), CM4AI (91.7%), CHORUS (84.5%)
- Add comprehensive About section explaining v5 generation and evaluation process

This enables GitHub Pages deployment of v5 HTML files.

Author: realmarcin

docs/d4d_examples.md

@@ -70,6 +70,30 @@ These datasheets were automatically synthesized using Claude Sonnet 4.5 with **d
 - [Download YAML](yaml_output/concatenated/claudecode/VOICE_d4d.yaml)
 - [Download Metadata](yaml_output/concatenated/claudecode/VOICE_d4d_metadata.yaml)
 
+## Latest: v5 Datasheets and Semantic Evaluations (December 2025)
+
+**NEW**: These are the latest datasheets (v5) with comprehensive semantic evaluation results.
+
+### AI-READI v5
+- [D4D Datasheet (Human-Readable HTML)](html_output/D4D_-_AI-READI_v5_human_readable.html) - Comprehensive project metadata
+- [Semantic Evaluation Report](html_output/D4D_-_AI-READI_v5_evaluation.html) - Rubric20 assessment (79/84, 94.0%)
+
+### CHORUS v5
+- [D4D Datasheet (Human-Readable HTML)](html_output/D4D_-_CHORUS_v5_human_readable.html) - Comprehensive project metadata
+- [Semantic Evaluation Report](html_output/D4D_-_CHORUS_v5_evaluation.html) - Rubric20 assessment (71/84, 84.5%)
+
+### CM4AI v5
+- [D4D Datasheet (Human-Readable HTML)](html_output/D4D_-_CM4AI_v5_human_readable.html) - Comprehensive project metadata
+- [Semantic Evaluation Report](html_output/D4D_-_CM4AI_v5_evaluation.html) - Rubric20 assessment (77/84, 91.7%)
+
+### VOICE v5
+- [D4D Datasheet (Human-Readable HTML)](html_output/D4D_-_VOICE_v5_human_readable.html) - Comprehensive project metadata
+- [Semantic Evaluation Report](html_output/D4D_-_VOICE_v5_evaluation.html) - Rubric20 assessment (81/84, 96.4%)
+
+**About v5 Datasheets**: Generated using Claude Sonnet 4.5 with deterministic settings (temperature=0.0, model: claude-sonnet-4-5-20250929) on December 20, 2025. Each datasheet has been evaluated using the Rubric20-Semantic framework, which assesses 20 questions across 4 categories (Structural Completeness, Metadata Quality, Technical Documentation, FAIRness & Accessibility) with semantic validation of correctness and consistency.
+
+**Average Score**: 77.0/84 (91.7%) across all 4 projects.
+
 ## Individual Dataset Datasheets
 
 These datasheets were created from specific dataset metadata sources:
@@ -134,6 +158,45 @@ See [DETERMINISM.md](https://github.com/bridge2ai/data-sheets-schema/blob/main/D
 ### Individual Dataset Datasheets
 The **Individual Dataset Datasheets** provide detailed metadata for specific datasets from each project's primary data repository (FAIRHub, Dataverse, PhysioNet). These focus on individual dataset instances rather than project-level metadata.
 
+### v5 Datasheets and Semantic Evaluations
+The **v5 Datasheets** represent the latest generation (December 2025) of comprehensive project metadata created using Claude Sonnet 4.5 with fully deterministic settings:
+
+**Generation Process:**
+1. Multiple project-related documents concatenated in reproducible order
+2. AI-powered extraction and synthesis using Claude Sonnet 4.5
+3. Temperature=0.0 for deterministic output
+4. Pinned model version (claude-sonnet-4-5-20250929) for reproducibility
+5. Validation against the LinkML D4D schema
+
+**Semantic Evaluation Framework:**
+Each v5 datasheet has been evaluated using the **Rubric20-Semantic** framework, which provides:
+
+- **20 Questions** across 4 categories:
+  1. Structural Completeness (max 24 points) - Schema field population and required elements
+  2. Metadata Quality (max 22 points) - Accuracy, specificity, and completeness of information
+  3. Technical Documentation (max 25 points) - Collection, preprocessing, and reproducibility details
+  4. FAIRness & Accessibility (max 13 points) - Findability, Accessibility, Interoperability, Reusability
+
+- **Semantic Analysis:**
+  - Correctness validation (DOI formats, grant IDs, RRID identifiers)
+  - Consistency checking (cross-field validation, logical coherence)
+  - Content accuracy assessment
+  - Evidence-based scoring with detailed justifications
+
+**Evaluation Results:**
+- **VOICE**: 81/84 points (96.4%) - Highest scoring project
+- **AI-READI**: 79/84 points (94.0%)
+- **CM4AI**: 77/84 points (91.7%)
+- **CHORUS**: 71/84 points (84.5%)
+- **Average**: 77.0/84 points (91.7%)
+
+Each evaluation report provides:
+- Category-level score breakdown
+- Question-by-question assessment with evidence
+- Semantic analysis findings (correctness and consistency checks)
+- Strengths, weaknesses, and recommendations
+- Detailed justifications for all scoring decisions
+
 ## Schema Information
 
 All datasheets conform to the [Datasheets for Datasets](https://arxiv.org/abs/1803.09010) framework by Gebru et al., implemented using the [Bridge2AI LinkML schema](https://github.com/bridge2ai/data-sheets-schema).