Searchability
The Second Pillar of Open Source Research
Searchability means that research must be fully machine-readable, structurally explicit, and parseable.
1. Machine-Readable Structure
Documents must be encoded in a structured canonical format (Pandoc JSON AST). All sections, tables, figures, citations, and equations must be explicitly defined.
2. Structured Objects
Tables and images must not merely appear visually — they must exist as structured data.
- Tables encoded natively or imported from CSV
- Figures generated from data where possible
- SVG preferred over raster images
- All objects assigned unique identifiers
3. JSON Descriptions
All tables, figures, and multimedia elements must include structured JSON descriptions proportional to their complexity. These descriptions enable:
- Text-to-speech systems
- Knowledge graph extraction
- Advanced indexing
- Computational scholarship
4. Object-Level Addressability
Each structural unit of a document must be individually identifiable and referenceable. This allows:
- Precise quotation
- Programmatic extraction
- Version comparison
- Deep scholarly analysis
Searchability ensures that research is not merely readable — it is computationally interpretable.