Step 2.2 · First Insights

Find the Gaps: Content Gap Analysis

Presence is easy — any tool can tell you what's in your text. The graph tells you what's conspicuously missing: the gaps between your topics, which are also your best source of new ideas.

By Dmitry Paranyushkin · Updated

Why Gaps, Not Keywords

Everything most tools compute is a presence operation: what’s connected, what’s frequent. The distinctive move of network analysis is reasoning about structured absence. When two clusters of ideas are each well-developed but the space between them is empty, that empty space is informative: it’s a question your discourse is silently asking.

In network science these are known as structural holes — and the people (or texts) that bridge them are consistently the source of the most novel information. (the theory)
An InfraNodus knowledge graph with a structural gap highlighted between two distant topical clusters
A structural gap between two well-developed clusters: both topics are present, but nothing in the text relates them yet.

Reading Your Gaps

Open the Insight panel on your Step 1.1 graph:

  1. InfraNodus detects the topical clusters that are distinct and distant, and highlights the gap between them on the graph.
  2. The panel names the two clusters that could be better connected.
  3. You’re invited to think of the bridge yourself — or let the built-in AI phrase a research question connecting them.

Try the manual route at least once: read the concepts on both sides of the gap and ask “what would a sentence containing both look like?” The question that feels awkward is usually the interesting one. Then compare with the AI’s version.

InfraNodus highlighting a structural gap between two topical clusters and generating a question that bridges them with the built-in AI
The AI phrases the gap as a research question — a prompt for the idea your discourse is silently asking for.

From Gap to Idea

A bridged gap is not just a question — it’s a content plan:

  • For thinkers and note-takers: each gap is a journal prompt or a note that doesn’t exist yet.
  • For writers: an article that bridges two of your well-developed themes is one only you can write — the graph guarantees it’s not the obvious take. This is also why gap-bridging content scores on informational gain in search (content gap analysis for SEO).
  • For researchers: gaps across a corpus of papers are literature-review gold: questions at the intersection of two literatures (research workflows).

Add the bridging idea back into the graph as a new statement and watch the structure change — the gap closes, and new gaps appear. That iterative loop is the heart of the next step.

Same Insight, in Your LLM

  • “What are the content gaps in my graph ‘[name]’?” (generate_content_gaps)
  • “Generate three research questions that bridge the biggest gaps.” (generate_research_questions — use useSeveralGaps for diversity, gapDepth to reach the less obvious gaps)
  • “Generate ideas that bridge the gaps and bring in outside perspectives.” (generate_research_ideas with shouldTranscend)