Entity Relationship Charts

← Back to Analyst's Notebook Learning Path

Master the art of making boxes and lines tell stories that would make detective shows jealous.

Building Effective Entity Relationship Charts (Or: How to Make Your Boss Think You're a Genius)

Entity relationship charts are the foundation of visual analysis in Analyst's Notebook. These charts allow you to represent people, organizations, events, locations, and other entities as icons and the relationships between them as links - essentially turning your investigation into something that looks like it belongs on a conspiracy theorist's wall, but with better graphics.

Entity Chart Truth #1

The moment you present your chart is the exact moment someone will ask, "But what about this connection you missed?" Be prepared to smile through the pain.

Key Skills Covered (That Will Make You Look Like a TV Detective)

• Creating different types of entities - From "Definitely Suspicious Person" to "Probably Innocent Bystander"

• Adding properties and attributes - Because knowing someone's favorite ice cream flavor might be important to the case (but probably isn't)

• Creating links between entities - Drawing lines that say "these two people definitely know each other" based on one phone call three years ago

• Defining relationship types - Going beyond "knows" to the more specific "owes money to" or "pretends not to recognize at parties"

• Organizing and arranging your chart - The art of making your tangled web of connections look intentional and insightful

• Using the palette - Finding that one specific icon type that's hiding in a submenu of a submenu

• Customizing entity appearances - Making your chart colorful enough to be interesting but not so colorful it induces seizures

Best Practices (That Everyone Agrees With But Few Actually Follow)

When creating entity relationship charts, consistency is key. Use standard entity types and link types to ensure your charts are easily understood by others - or at least as easily understood as a complex network of criminal relationships can be.

• Use appropriate entity types - A person is a person, not a building. This seems obvious until you're 14 hours into chart creation.

• Label links clearly - "Associated with" is the analytical equivalent of shrugging your shoulders. Be specific!

• Include relevant attributes - But remember, not everything needs to be on the chart. No one needs to see a suspect's entire Netflix viewing history.

• Arrange entities logically - Your chart should not resemble a Jackson Pollock painting unless the case involves abstract expressionists.

• Use consistent styling - Pick a color scheme and stick with it. This isn't the time to express your artistic side.

The Entity Chart Paradox

A chart with too few entities looks incomplete. A chart with too many entities looks incomprehensible. The perfect chart exists in a quantum state between these two extremes that can never actually be achieved.

Common Entity Chart Disasters (And How to Recover)

• The "Everything Is Connected to Everything" Syndrome - When your chart has so many crossing lines it resembles a ball of yarn after a cat attack

• The "Mystery Icon" Problem - When you use a custom icon that made sense at 2 AM but is completely baffling in the morning

• The "What Does This Link Mean Again?" Dilemma - When you forgot to label a critical connection and now have to pretend you remember what it represents

• The "My Chart Is Too Big for the Projector" Crisis - When you have to zoom out so far that all your careful labeling becomes microscopic

Additional Resources

-

Entity Relationship Templates (For the Creatively Challenged)

-

Chart Style Guide (Making Your Analysis Pretty AND Functional)

Next Topics

-

Timeline Visualization (Time Is A Flat Circle, But Your Chart Shouldn't Be)

-

Social Network Analysis (Finding the Popular Kids in Criminal Networks)

What Are Entity-Relationship Charts?

Entity-relationship (ER) charts are a visual analytical tool used to map and analyse the connections between people, organisations, locations, events, and other entities relevant to an intelligence investigation. Unlike generic association charts, ER charts in an intelligence context incorporate analytical weighting, temporal context, and evidentiary standards.

• Beyond Simple Association: An ER chart does more than show who knows whom. It incorporates the nature, strength, direction, and duration of relationships, as well as the evidentiary basis for each connection.
• Dynamic Analysis: ER charts are not static products. As new information becomes available, entities and relationships are added, modified, or removed. The evolving chart itself becomes a record of the analytical process.
• Hypothesis Testing: ER charts are a tool for testing hypotheses about structure and relationships. Does the chart suggest a hierarchical organisation or a distributed network? Are key individuals connected to multiple otherwise separate clusters?

Building an Entity-Relationship Chart

Effective ER charts follow a systematic construction process:

1. Entity Identification: Identify all relevant entities from available source material. This includes not only obvious entities (named individuals, known organisations) but also latent entities that emerge from the data (common phone numbers, shared addresses, recurring patterns).
2. Relationship Extraction: For each pair of entities, determine whether a relationship exists and, if so, what type. Relationship types include direct communication, financial transaction, shared location, common membership, and familial association.
3. Strength and Confidence Assessment: Not all relationships are equally significant. A single phone call between two numbers is weaker evidence of a relationship than 200 calls over six months. Assign relationship strength and analytical confidence based on the available evidence.
4. Temporal Placement: Place entities and relationships in time where possible. Knowing when a relationship was active is often as important as knowing it existed. A relationship that ended before the investigation period may be irrelevant; one that began during the investigation period may be highly significant.
5. Visual Arrangement: Arrange entities on the chart to reveal patterns. Place central or high-value entities near the centre, cluster related entities together, and use proximity to indicate relationship strength.

Key Patterns to Identify

ER charts are particularly effective at revealing several recurring structural patterns:

• The Star Pattern: A central entity connected to many peripheral entities with few connections among themselves. This pattern suggests a hub-and-spoke structure, common in drug trafficking networks and intelligence rings where cell members communicate through a single handler.
• The Chain Pattern: Entities connected in a linear sequence - A connects to B, B connects to C, C connects to D. Chains often indicate hierarchical communication paths or supply chains. Identifying chain breakpoints reveals critical nodes for disruption.
• The Cluster Pattern: Multiple entities with dense interconnections forming a distinct group, with few connections to entities outside the group. Clusters indicate cohesive sub-groups - teams, cells, or departments within larger organisations.
• The Bridge Pattern: An entity that connects two otherwise separate clusters. Bridge entities are analytically significant because they control information flow between groups and are often the most effective targets for network disruption.
• The Isolate Pattern: An entity with no connections in the chart. Isolates may be irrelevant to the investigation, may represent entities whose connections haven't yet been discovered, or may be deliberately compartmented operatives.

Analytical Pitfalls

Common mistakes in ER chart construction and interpretation include:

• Over-Inclusion: Including every entity and relationship without analytical triage produces an unreadable chart. Apply the "so what?" test - if removing an entity wouldn't change the analytical picture, consider whether it needs to be in the chart.
• Assuming Absence Means Absence: The absence of a visible connection does not mean no connection exists. It may mean the connection operates through channels not captured in available data - encrypted communications, in-person meetings, or cutout intermediaries.
• Presenting Correlation as Causation: Two entities that both connect to a third entity are not necessarily connected to each other. A correlation in the chart may reflect a common activity rather than a direct relationship.