A discourse graph

Resilient Data Futures

The SciOS Resilient Data Futures whitepaper argues that research data loss is architectural; that the accumulated loss is an institutional liability; and that the solution to data loss hedges the liability while producing the data substrate AI-ready institutions are looking for. We publish that argument as a discourse graph — every claim, evidence item, question, method, and source is its own addressable node. The graph is canonical. The paper is one rendered view over it.

What is a discourse graph?Explore the graph Read narratives

Five node types

The graph in numbers

Question

Open lines the work addresses. Each Question gathers Claims that respond to it.

Claim

Assertions. Each Claim addresses a Question, may be supported or opposed by Evidence and other Claims.

Evidence

112

Specific empirical observations grounded in a Source. Each item supports or opposes a Claim.

Method

Analytical instruments — taxonomies, formulas, frameworks — that Claims invoke.

Source

134

References: papers, datasets, archived records. Provenance backstop for Evidence.

Three ways in

Pick your projection

Topology

The whole graph at a glance. Nodes coloured by type, edges by relation. Filter, cluster, traverse.

Narratives

The original whitepaper that seeded the graph, plus narratives composed from the graph's nodes for different audiences and framings.

Per-node browse

Every node addressable on its own. Backlinks computed at build time so the graph is bidirectionally traversable.