research systems

methods for knowledge and evaluation Hélène Draux

This site presents methods for describing, structuring, and evaluating research systems and their outputs. It focuses on how knowledge can be represented, analysed, and tested, including under conditions of increasing automation.

The emphasis is on clarity, structure, and traceable reasoning rather than completeness.

I work at the intersection of bibliometrics, knowledge infrastructure, and AI-assisted research. This site documents that work — methods developed over several years, alongside tools and writing that sit outside the formal structure.

---

Methods

Conceptual approaches for representing and analysing the structure of research systems.

• AI metascientist
• Journalscape
• Writing research publications

---

Frameworks

Structured applications of methods in defined analytical contexts.

• GRID+
• BiblioFlow
• DimQuery

---

Tools

Lightweight implementations supporting specific analytical tasks.

• AVA
• URL to DSL

---

Projects

Things built or written outside the formal structure — tools for personal workflows, visual and analytical experiments, work in progress.

• Loom
• Extending Tufte: the bumped stacked bar chart
• Small multiples
• Claude Projects as semi-structured workflows
• Semantic layer over Vinted (work in progress)
• Local voice assistant for kids

---

Perspective

The work is situated within bibliometrics and research systems analysis, with a growing focus on evaluation under conditions of increasing automation.

Methods are documented at a level that supports interpretation and reuse, while full operational detail is not always public.

---

Structure

Methods → Frameworks → Tools Conceptual → Applied → Operational

Methods inform frameworks. Frameworks organise tools.