How to Visualise Your Site as a Knowledge Graph

The Graph Explorer in QueryBurst renders the site's extracted knowledge as a navigable graph — hop-by-hop from any seed entity, showing connected entities, relationship predicates (e.g. "founded", "leads", "located in"), degree distance, and centrality scoring. Use it to see how the site's entities connect, which are most central, and where relationship gaps weaken topical coherence.

How to Access the QueryBurst Graph Explorer

  1. Open a site in QueryBurst
  2. Click Site Intelligence in the sidebar
  3. Select the Graph Explorer subtab

Understanding the Graph Explorer Interface

Seed Entity

The graph explorer starts from a seed entity — by default, the entity that appears on the most pages. Everything is measured relative to this anchor point.

To change the seed entity, click the search icon next to the entity name and search for a different starting point.

Path Trail

The horizontal breadcrumb trail at the top tracks your journey through the graph. Each step shows:

  • The entity name with a colour-coded dot (by type)
  • The predicate (relationship label) that connected you
  • The degree distance from the seed entity

Click any step in the trail to jump back to that point.

Connections Panel (Left)

Shows all outgoing connections from the current entity, grouped by type:

  • Type tiles — When multiple types exist, click a type tile to see entities of that type
  • Entity list — Each connected entity shows its name, relationship predicate, a rank bar, and degree distance badge
  • Rank bar — Relative importance based on centrality scoring from the seed entity
  • Return arrow — Appears on entities that connect directly back to the seed

Click any entity in the list to drill into it and extend your path.

Minimap (Right)

A 2D force-directed graph showing:

  • Current entity — Highlighted in white
  • Seed entity — Highlighted in amber
  • Trail nodes — Coloured by entity type along your path
  • Return path — Faded amber nodes showing the shortest route back to the seed

Click nodes in the minimap to navigate to them (if they're direct connections of the current node).

Shortest Path

When you've navigated away from the seed, a shortest path button appears. Toggle it to see the shortest route back, displayed as a horizontal chain above the path trail.

Replay

If your trail has two or more hops, a Replay journey button lets you replay your walk step by step, auto-advancing every 1.2 seconds.

Key Metrics

MetricDescription
Degree distanceNumber of hops from the seed entity (BFS shortest path)
RankCentrality score relative to the seed — higher means more connected via important paths
ConnectionsNumber of outgoing relationships from the current entity

Interpreting Results

Healthy Signs

  • Well-connected core entities — Your main brand entities have many connections across types
  • Low degree distances — Most entities are reachable within 2–3 hops from the seed
  • Strong rank scores — Important entities rank high relative to the seed

Warning Signs

  • Isolated entities — Entities with no outgoing connections or very high degree distances
  • Disconnected clusters — Groups of entities that don't connect back to the main graph
  • Missing predicates — Key relationships you expect to see aren't present in the graph

Tips

  1. Start from your brand entity — Set your main brand as the seed to see how everything connects to it
  2. Look for missing connections — If two entities should be related but aren't connected, the content may need stronger linking language
  3. Compare degree distances — Entities that are conceptually close but have high degree distances may need bridging content
  4. Use replay for presentations — The step-by-step replay is useful for demonstrating knowledge structure to stakeholders

Technical Details

The knowledge graph is built from subject–predicate–object triples extracted during site analysis. Each edge represents a relationship like "offers", "is located in", or "was founded by". Nodes must appear on at least one page to be included.

Centrality is computed with the seed entity as the anchor, measuring each node's importance relative to the seed rather than globally.

  • Knowledge Graph — Entity profiles table and summary stats
  • Entity Profile — Deep dive into a single entity's statements and pages
  • Link Explorer — Similar walker interface for the internal link graph