What Is a Link Node Chart?
A link node chart shows relationships (”links”) between objects (”nodes”).
Workstation has robust tools for automatically categorizing objects, identifying relationships, and visualizing them in an easy to explore way. For intelligence and investigations, the platform can show how people, places, vehicles, phones, accounts and events connect.
If you’ve ever seen a crime drama showing a bulletin board with colored string connecting things, this is a link-node chart!
Why Use a Link Node Chart?
Most cases don't stall for lack of data. They stall because the data lives in pieces: arrest records in one system, phone records in another, bank records somewhere else. A table shows you rows. A chart shows you relationships, and that changes what you can see.
See connections a table hides. Two nodes may share multiple connections that are hidden in different tables.
Find who matters. Some people and things sit at the center, or connect separate groups.
Spot the gaps. A node with no links, or a link you expected but can't find, tells you where to look next.
Explain the case. A picture is faster to brief than a spreadsheet.
When Should You Use One?
A good rule of thumb: if you're holding names and numbers in your head, or flipping between tabs to check whether two things are connected, it's time for a chart.
The same name, number, plate or account shows up in more than one system.
You suspect two cases are related but can't show it yet.
You're working a network: a gang, a fraud ring, a racketeering case, a supply chain.
You need to show your boss or a prosecutor how the pieces fit.
Skip it when you have a handful of records and a simple list answers the question. A chart earns its place when the connections are the question.
How Do You Read One?
Every chart is built from the same few ingredients, and six terms cover most of what you’ll see:
Term | Plain meaning | What it tells you |
|---|---|---|
Node | One thing: a person, phone, vehicle, account or place | Who and what is in the case |
Link | A relationship between two nodes | How they connect |
Cluster | A tight group of nodes | A crew, a family, or a business |
Hub | A node with many links (high "degree") | Who or what is central |
Bridge | A node that connects two groups (high "betweenness"). Also called a broker or gateway. | Who passes information, money or goods between groups |
Outlier | A node with few or no links | Who might be missing a connection |
Start with hubs and bridges. A hub shows you who is busiest. A bridge shows you who is holding two groups together, and that is often the more interesting lead.
What Can You Find?
The same chart answers very different questions depending on what you put into it. Here are four, with what stands out in each. These are illustrative, and there are many more.
Question | What you chart | What stands out |
|---|---|---|
Are these thefts connected? | Reports from several departments | The same tow company, address and phone across cases |
Where did the money go? | Accounts and transfers | One account that receives from many and pays out to a few |
Who owns what? | Companies, directors, registered addresses | Several "unrelated" firms sharing a director and an address |
Who links two groups? | Members, phones, shared addresses | One person with links into both clusters |
Take that last one. In a racketeering case, you chart the members, their phones and the addresses they share. Two tight clusters sit apart from each other, and one person has links into both. That person is the bridge. They may not be the biggest name on the chart, but they are the connection between the groups, which makes them a strong place to start.
How Has Link Analysis Evolved?
The idea is old. Investigators have pinned photos to a wall and run string between them for generations. What has changed is how much of the work the tool does for you.
Stage | What it is | What still takes work |
|---|---|---|
1. Corkboard | Photos, notes and string on a wall | Every link is drawn by hand, and it's hard to share |
2. Digital canvas (e.g., i2 Analyst's Notebook) | Software to draw and arrange the chart on screen | The analyst has to build and remember to update it when it changes |
3. Shared, centralized source (e.g., Palantir) | One system the whole team works from | Data often has to be loaded and transformed first |
4. AI-led (e.g., Workstation) | AI reads the records, categorizes them into types, proposes the ontology, derives the connections, and draws the chart | The analyst reviews, corrects and directs |
Each previous stage kept the analyst's judgment and moved more of the setup work elsewhere. In the AI-led stage, the AI automatically figures out the nodes and edges (including ones that may not be obvious), thus first draft of the chart takes minutes instead of days, leaving more of the analyst's time for reviewing and reading the picture.
How Do You Build One?
The steps to mapping relationships are the same, but the difference is how many of them you do yourself.
Start with a question. "Who connects these two groups?" beats "show me everything."
List the types of things you care about. These are usually nouns. People, phones, vehicles, accounts, addresses, events.
Gather the records. Pull from every system that touches the question.
Merge duplicates. "J. Smith," "John Smith" and "Johnny S." may be one node.
Add the links. These are usually verbs. Label each one: called, owns, drove, paid, lives at.
Look for hubs, bridges, clusters and outliers. This is where the value surfaces.
Check the sources, then share it. Every link should trace back to a verifiable record.
In Workstation, you upload records or connect your sources and ask the AI for a link node chart. It handles steps 2 through 5, and the analyst reviews the result. Items are sorted into types with their own icons and colors, and built-in network analysis highlights the hubs, bridges and clusters. You can turn sources on and off to see what each one adds. Charts are versioned and shareable, and links trace back to the source data. To add or change a node, tell the AI what you want.
Link Node Chart FAQ
What is a link node chart used for?
Showing how people, places, things and events connect, so you can find key players, hidden links and gaps in a case or a dataset.Is a link node chart the same as a network graph?
They're closely related. "Link node chart" is the common term in intelligence and investigations. "Network graph" is the common term in data science.How is an AI-assisted chart different from i2 Analyst's Notebook?
Analyst's Notebook is a digital canvas: the analyst builds and arranges the chart. In an AI-assisted tool, the AI identifies what should go on the chart, classifies the links and nodes into types, organizes the chart into communities, and allows you to interact with the chart without dragging lines or bubbles around.Can I trust a chart the AI built?
Treat it as a draft. Each link should trace back to a source record, and the analyst decides what holds up.

Workstation
Team



