One of the most important aspects of data is understanding how it relates to itself or other data. However, the discovery of relationships in data becomes increasingly difficult if you do not have a way to visualize relational data. So, before you start doing this the old-school way with pins, strings, and a bulletin board, let’s look at the new-school way with link charts in ArcGIS AllSource.

A link chart visualizes relationships between entities in data and is a complementary view to the map. Link charts consist of two main components, nodes and links. Nodes represent entities that are people, places, and things, for example, a name, phone number, or an IP address. Links are the lines connecting nodes and are the visual representation of related/connected entities.

Link charts are created using attributes in your data depending on what you want to visualize and how your data is modeled, without requiring complicated schemas or predefined ontologies. The data you use to create a link chart can be contained within a single feature layer, multiple feature layers, or in a stand-alone table. In this part of the blog, you are going to learn how to create a link chart from a single feature layer and multiple features layers, however, the process is the same for stand-alone tables.

Build a link chart from a single feature layer

The link chart above shows a typical organizational chart for a fictional military organization. All the units are entities, and the links show the hierarchy of the organization which is the relationship between these units. You will create the same link chart using data from a single feature layer. However, before you begin, let’s examine the data to gain a better understanding of how it’s organized. The Units feature layer has each units’ name stored in the Unique_Name field. Notice that there are some duplicates in the Higher_Unit field because these higher units have more than one unit under their command.

The first step to create a link chart is to create an empty link chart. To create an empty link chart in a project, complete the following steps:

1. On the Analysis tab, in the Link Analysis group, click New Link Chart. The Link Chart view appears. Additionally, the chart is listed in the Contents pane.
2. Optionally, rename the link chart on the Link Chart properties dialog box.

Next, you’ll create entity types from a field in your feature layer. To create an entity type, complete the following steps:

1. On the Link Chart contextual tab, in the Add group, click Entity Type. The Add new entity type from layer dialog box appears.
2. In the Choose Layer drop-down menu, choose the Units layer.
3. In the Choose Field(s) for entities drop-down menu, choose Unique_Name. Choosing the Unique_Name field creates an entity for each unique value in this field.
4. Check the Visually collapse duplicate entities box. This will collapse duplicate fields so that you don’t have multiple entities of the same unit showing up in the link chart.

Next, you will create the relationship type to define how the entities are connected in the link chart. To create a relationship type, complete the following steps:

1. On the Link Chart contextual tab, in the Add group, click Relationship Type. The Add new relationship type between entities dialog box appears.
2. In the Choose source entities drop-down menu, choose Unit.
3. In the Choose target entities drop-down menu, choose Unit.
4. In the Enter name for relationship type text box, name your relationship Unit to Higher Unit.
5. In the Choose key type drop-down menu, choose Entities.
6. In the Source entity key field drop-down menu, choose Unique_Name.
7. In the Target key field drop-down menu, choose Higher_Unit.
8. Click OK.

When you create a relationship using the key type Entities, you are using attributes from the source and target entity types to link the entities based on matching attribute values. In this example, you’ve created a relationship that shows the link between units and their commanding units.

Build a link chart from multiple feature layers

Now, you will use two feature layers to create a link chart that shows the relationships between movies and actors. The attribute tables for these two feature layers are shown below.

The Movies attribute table (left) contains a record for each movie and the actors that were in that movie. The actors’ identity is stored in this attribute table by a unique Actor_ID field value (this is representative of other unique IDs such as social security numbers, cross reference numbers, prisoner IDs, and so on). The attribute table on the right contains the same list of Actor IDs, in a similar field. For example, in the Movies attribute table the selected record for the movie Predator has an Actor_ID of 216, which corresponds with the Actor_ID field value of 216 in the Actors attribute table; this Actor_ID happens to be Arnold Schwarzenegger.

Understanding how the data is organized and what you want to visualize in the link chart will help you create the appropriate entity and relationship types. In this example, you will create two entity types: one for movies and another for actors.

To create the Movies entity type, complete the following steps:

1. Create an empty link chart.
2. On the Link Chart contextual tab, in the Add group, click Entity Type. The Add new entity type from layer dialog box appears.
3. In the Choose Layer drop-down menu, choose the Movies layer.
4. In the Choose field(s) for entities drop-down menu, choose the Movie field.
5. In the Enter name for entity type text box, name your entity type Movies.
6. Check the Visually collapse duplicate entities check box because the Movies field contains multiple entries per movie.

To create the Actors entity type, complete the following steps:

1. On the Link Chart contextual tab, in the Add group, click Entity Type. The Add new entity type from layer dialog box appears.
2. In the Choose layer drop-down menu, choose the Actors layer.
3. In the Choose field(s) for entities drop-down menu, choose the Name field.
4. In the Enter name for entity type text box, name your entity type Actors.
5. Check the Visually collapse duplicate entries check box.
6. Click OK.

Once the entity types are created, you can define how the entities are connected through relationship types. As discussed earlier, the feature layers associated with these entity types have a common key field for actors, the Actor_ID field. By setting key fields to this common field, links are generated between entities when you create the relationship type. Note that in this case, the linking field happens to be named the same, however, that is not a requirement. Next, you want to visualize the cast of each movie, so you’ll define the relationship type.

To define the relationship type, complete the following steps:

1. On the Link Chart contextual tab, in the Add group, click Relationship Type. The Add new relationship type between entities dialog box appears.
2. In the Choose source entities drop-down menu, choose Actors.
3. In the Choose target entities drop-down menu, choose Movies.
4. In the Enter name for relationship type text box, type Actors to Movies.
5. In the Choose key type drop-down menu, choose Entities.
6. In the Source entity key field drop-down menu, choose Actor_ID.
7. In the Target entity key field drop-down menu, choose Actor_ID.
8. Click OK.

Now that the relationship type has been created, it’s easy to see in the link chart (below) which movies Arnold Schwarzenegger performed in and his associated cast members.

Analyze the link chart

Link charts enable you to visualize relationships and add further context to your data and map. However, in addition to visualizing relationships in your data, link charts also have analysis tools to aid you in your work. In Part 2 of this blog, you will learn about the Centrality, Cluster, Neighborhood, and Path Link Analysis tools available in ArcGIS AllSource.

Tips and tricks

• Entity types can be created in any order.
• If your feature layers do not have fields that contain linking values between entity types, you can skip setting the key fields by setting the key type to None when you create the relationship type. After the relationship type is created, you can manually draw links between entities using the Add Link tool.
• If your feature layers have fields that contain linking values between entity types, but not all records have values, set the key type to Entities so that a link is created for the entities that do have linking values. For entities that don’t have a linking value you can manually draw links using the Add Link tool.
• The Change Layout drop-down menu consists of three layouts—Hierarchic, Clustered, and Organic—that rearrange entities and relationships according to the new layout. Some layout types display a given set of data better than others so you may need to try different layout types to see which works best for your data and use case.

Related resources

Visit the ArcGIS AllSource website.

To learn more, see the link chart help documentation.

Continue to Part 2 to learn about link analysis tools.

Information courtesy of IMDb (https://www.imdb.com).
Used with permission.

Card photo by Volodymyr Hryshchenko on Unsplash

Banner photo by Planet Volumes on Unsplash

About the authors

Stephen Kredel

Stephen Kredel is a Product Engineer on the ArcGIS AllSource team. He joined Esri in 2022 and focuses on technical writing and help documentation. He has four years experience as a Geospatial Intelligence Imagery Analyst in the U.S. Army and earned his B.A. in history from the University of Redlands.

Eric Linz

Eric Linz is a Senior Product Engineer on the ArcGIS AllSource team. Eric has been with Esri for over 30 years, during which time he has supported the defense and intelligence communities through software development and data production. His educational background is in Geography.

Tony Angell

Tony is a product engineer on the Intelligence Solution team at Esri. He has been with Esri since 2019 and a former Marine with multiple years of support to the SOF community.