Table of contents
- Overview of Communication Visualizer
- Accessing Communication Visualizer
- How does Communication Visualizer work?
- Graph components
- Using Communication Visualizer
- Communication Visualizer for Common Workflows
Overview of Communication Visualizer
Communication Visualizer is an interactive and filterable network graph of select email communications that helps you gather insights into communication patterns. Understand who was communicating with whom, with what frequency, and how those communication patterns evolved over time, to filter down to relationships of interest and review the emails involved.
Communication Visualizer is best leveraged in investigations and workflows throughout the discovery process to identify key communicators, understand communication patterns for specific persons of interest, and identify anomalies in those patterns.
Communication Visualizer is directly integrated with Everlaw’s Data Visualizer and allows you to generate visualizations for searches containing up to 1 million emails. Learn more about how Communication Visualizer works.
After watching the introduction video above, we recommend exploring Communication Visualizer with a subset of emails in a given project and diving into some common workflows. You can learn more about graph components here and see a comprehensive overview of using Communication Visualizer here.
Accessing Communication Visualizer
Communication Visualizer is directly integrated with Data Visualizer and can be accessed similarly from a search page, results table, or through the Document Analytics icon in the top navigation toolbar. Note that users subject to document access management restrictions will only be able to visualize emails they can access.
To access Communication Visualizer from the search page, select “Open data visualizer” after building your initial search.
From a results table, select the “Visualize” option from the top toolbar.
From the top navigation toolbar, click the Document Analytics icon and then select Data Visualizer from the dropdown menu.
Once in Everlaw’s Data Visualizer, select the “Communication” tab on the left hand side of the screen to generate a Communication Visualizer graph.
How does Communication Visualizer work?
Communication Visualizer automatically analyzes the contact names, email addresses, and domains of emails in a given targeted search to generate a visualization of email communication patterns.
To be visualized, emails must have values for the following metadata fields: Date sent, From, and one or more values for To/Cc/Bcc. In addition, emails must have a valid date sent value between January 1, 1971 and present-day. Any emails that do not have the appropriate values for these listed fields will not be displayed on visualizations generated by Communication Visualizer.
Graph Components
Communication Visualizer generates graphs with three main components:
- Nodes (the circles)
- Edges (the lines connecting nodes)
- the “Others” node (the quarter circle in the top left corner of the visualization)
Nodes
Nodes in Communication Visualizer represent a contact name, email address, or domain. Each one of these options can be selected as the node type for a given visualization.
- The size of a given node indicates the volume of emails (i.e., the number of sent and received emails) associated with the node. Hover over a given node to see the breakdown of the number of emails associated with the node.
- By default, nodes in the visualization will appear closer to nodes that they either (a) communicate with directly or (b) share mutual communicators with. For example, people on the same team may appear closer together in a visualization.
Edges
Edges represent the volume and directionality of the emails exchanged between nodes.
- The thickness of an edge indicates the number of emails sent between the two nodes. Hover over an edge to see the number of emails sent using the To, CC, and BCC fields.
- The location and direction of the arrow on each edge indicates the proportional directionality or flow of communication between those nodes. An arrow near one end of an edge indicates that most communications flow in that direction, while an arrow near the middle indicates a similar volume of emails from each node. In addition, the arrow’s position along the edge indicates the actual proportion of email communications that are being sent from either node.
“Others” node
The "Others" node captures all communications between nodes displayed in the visualization and other nodes in the document set that are not currently displayed in the visualization. This node gives you visibility into all communications involving the top, displayed nodes. Note that the “Others” node does not include communications between nodes that are not displayed in the visualization.
For example, let’s assume the following:
- A is displayed as one of top nodes in the visualization
- B, C, and D are not one of the top nodes in the dataset so they are not displayed in the visualization
- A communicates with B and C
- A does not communicate with D
- D communicates with B
Here, the “Others” node will include any emails between A and B as well A and C because A is displayed in the visualization. However, the “Others node” will not include communications between B and C or B and D because these nodes are not displayed in the visualization.
Using Communication Visualizer
Basic Navigation
- Click and drag: click and drag on a visualization to pan around the page
- Hover over details: hover over nodes and edges to gather more information about the number of emails (sent or received) associated with a given graph component.
- Zoom in and out: zoom in and out of visualizations by clicking on the magnifying glass icons with the plus and minus, scrolling with a mouse wheel, or using pinch-to-zoom on a laptop.
- Undo and redo: undo and redo the last selection made on the visualization by clicking the left and right arrow icons in the toolbar. Note that the undo and redo behavior does not apply to actions taken in Explore mode.
- Reset view: After panning around the page and zooming in/out of the visualization, you can reset your initial view of the visualization by clicking the “Reset view” link in the far right corner of the Communication Visualizer toolbar.
Edit visualization display options
Change default node type and top number of nodes displayed
By default, Communication Visualizer will display the top 10 name nodes with the highest number of emails sent and received within the underlying search.
Depending on what you would like to focus on in your visualization, you can change the node type and number of top nodes to display in the visualization. For example, if you are interested in gathering insight on the use of personal email addresses (e.g., @google.com, @yahoo.com, etc.), we recommend you change the node type to domains.
To change the node type and the number of top nodes to display in the visualization, click “Edit” from the toolbar to open the “Edit visualization display options” menu.
You can select the node type to display (i.e., Names, Domains, or Emails) and enter the number of top nodes to display, then click the “Recalculate” button to regenerate your visualization based on these selections. You can currently display up to 25 nodes in a visualization when displaying name and email node types, and up to 20 nodes when selecting the domain node type.
Display specific nodes
You can also edit the visualization to select specific nodes to display with or without a specified number of their top neighboring nodes. Selecting specific nodes to display with top neighboring nodes can be especially helpful if you are interested in examining email communication patterns between a specific set of name, email, or domain type nodes.
Select your desired node type and the radio button next to “Display specific [selected node type].”. Then, click into the search bar to choose specific nodes to display or type in a specific value and select from the dropdown menu.
Once you have specified nodes to display, you have two options:
- (a) display top x neighbors for each of the selected nodes or
- (b) don’t display neighbors.
Top neighboring nodes or neighbors refers to the top nodes (i.e., highest numbers of email communications) connected by edges to a specific node.
If option (a) display top x neighbors is selected, Communication Visualizer will display the top X nodes that each of the specified nodes have communicated with most (the "neighbors"). In this option, the total number of nodes displayed must be less than or equal to the max number of nodes for the selected node type (i.e., up to 25 total nodes for name and email node type visualizations, up to 20 for domain type visualizations).
For example, let’s say you know that Ashley and Brian send emails to each other often and you want to know who else these specific individuals emailed frequently. If we select Ashley and Brian as the two specific name nodes to display, we can select up to 11 top neighboring nodes to display for each of the two specified names (i.e., 11 top neighbors for Ashley and 11 top neighbors for Brian)because the total number of name nodes displayed would be less than the 25 node maximum for names (i.e., 2 specified name nodes + 11 top neighbors for Ashley + 11 top neighbors for Brian = 24 total nodes). In contrast, if we instead selected two specific domain nodes to display, we can select up to 9 top neighboring nodes to display for each of the two specified domains because the total number of domain nodes displayed would be less than the 20 node maximum for domains (i.e., 2 specified domain nodes + 9 top neighbors for Domain A + 9 top neighbors for Domain B = 20 total nodes).
Node type: Names (25 node max)
| A: # specified names | B: Max # top neighbors |
C: Resulting # nodes C = A * (B + 1) |
| 1 | 24 | 25 |
| 2 | 11 | 24 |
| 3 | 7 | 24 |
| 4 | 5 | 24 |
| 5 | 4 | 25 |
Note that if there is any overlap between the top neighbors of the specified nodes, the resulting number of nodes in the graph may be fewer than what column C indicates above.
If option (b) don’t display neighbors is selected instead, no additional nodes will be displayed, and only email communications exchanged strictly between the specified nodes (and no one else) will be represented in the visualization. Please note that if there are no email communications exchanged strictly between the specified nodes, the visualization will be empty.
Date filter
Leverage the date filter at the bottom of the page to narrow the date range of the email data displayed. You can expand the date filter at the bottom of the screen then pan around and zoom in/out. Click and drag to capture a specific date range or type in the relevant date range in the from/to fields to understand how communication patterns evolve over time.
- The dark bars on the date filter represent communications between the nodes displayed in the visualization. These include communications between the nodes displayed in the visualization and the “Others” node.
- The light bars represent communications that are in the underlying search but are not displayed in the visualization. These include communications within the "Others" node.
Interaction modes
Communication Visualizer has two interaction modes: Selection mode and Explore mode. In Selection mode, every click will add or remove documents from your current selection. In Explore mode, you can freely click on nodes and edges to view more details without affecting your current selection.
Selection mode
By default, Communication Visualizer will open in Selection mode. Once a node and/or edge on the visualization has been clicked, details about your overall selection will appear in the side menu.
You can also select nodes from the “Search [node type] to select” dropdown menu in the toolbar. This dropdown menu is especially helpful to quickly see all the nodes displayed in the visualization, sorted by decreasing communication volume.
You can modify your selection by clicking on additional nodes or edges, selecting more nodes from the dropdown menu in the toolbar, or by interacting with the checkboxes within the side menu that appears after selection.
Selecting nodes
There are some visual cues that indicate whether a node has been fully selected (all edges associated with the selected node are selected), partially selected (at least one edge associated with selected node is selected), or not selected (no edges are selected).
For example, when you first click on a node, you will see that the selected node and all connected edges, including the edge connected to the “Others” node, will be highlighted in blue.
Similar to clicking on nodes in the visualization, nodes selected from the toolbar dropdown will be highlighted in blue because they are fully selected (i.e., all edges associated with these nodes are selected). Fully selected nodes will be displayed with an orange background in this dropdown.
Adding filters
After making selections in a given visualization, you can filter documents by selecting “Add filter” in the side menu. After the filter is applied, the visualization will regenerate to reflect the newly filtered set of emails. Applied filters will be represented at the top of the screen with information about the selection. Remove applied filters by clicking the red x.
The image above displays the top 10 name node type visualization name node for Jeff Dasovich has been selected. Once “Add Filter” is clicked, a representation of the filter will appear at the top of the screen. The screenshot below shows an example of what a filter added from Communication Visualizer looks like.
This filter shows a preview of the selection. In this example, communications between Jeff Dasovich and 6 neighbors were selected, and there are 6 additional nodes in the selection. On hover, the full list of selected nodes and their respective number of selected neighbors is displayed.
Explore mode
Enter Explore mode to learn more about the displayed nodes and edges in the visualization and pull up granular details without impacting your current selection. To enter and exit Explore mode, click the Explore icon in Communication Visualizer’s toolbar.
Once in Explore mode, select a node or edge to see information about how many email communications were sent to and from the selected node or edge. The node or edge that is currently being “explored” will be displayed with an orange highlight, regardless of whether or not it was previously selected in Selection mode.
After taking a closer look at a node or edge, you can update your current selection from Explore mode by clicking the relevant button at the bottom of the Explore mode side menu. The Explore mode side menu will indicate what the pending changes to the selection would be if the “Update current selection” button is selected .
For example, in the image above, we previously selected the node for Jeff Dasovich in Selection mode (as indicated by the blue highlight), then entered Explore mode to look through any additional nodes or edges to add/remove/update our current selection. After clicking on the edge between James D Steffes and Richard Shapiro (as indicated by the edge being highlighted orange) to see the breakdown of email communications, we can add this to our current selection by clicking the “Add to current selection” button at the bottom of Explore mode’s side menu. When returning to Selection mode (i.e., deselect the Explore icon in the toolbar) in this example, we will see the edge between James D Steffes and Richard Shapiro highlighted in blue.
Communication Visualizer for common workflows
Communication Visualizer can be leveraged for a variety of workflows throughout investigations and the discovery lifecycle. Although this section provides some common workflows to get you started, you can use any search containing email communications to generate a graph with Communication Visualizer.
Identify key communicators within a dataset
In this scenario, you have just uploaded email data to Everlaw and want a high-level overview of the key communicators in your dataset.
- Click the Document Analytics icon and select “Data Visualizer” from the dropdown menu.
- Navigate to the “Communication” tab in the left-hand side of the Data Visualizer page.
- In the resulting visualization by default, you will see the top 10 individuals (i.e., Name nodes) that frequently communicate within the underlying search. You can increase the number of name nodes displayed to identify additional key communicators within the underlying search.
Note: If your project contains more than 1 million emails, you will need to narrow your search (e.g. use filters in Data Visualizer or refine search then visualize).
Identify key communicators discussing a given topic
In this scenario, you have already identified a narrowed set of emails involving potentially interesting topics (e.g., created a “Content” search or found results from clustering) and want to find out who is communicating about these topics.
- Create a search for emails containing specific content or clusters of emails pertaining to relevant terms.
- Select “Visualize” from the results table (or “Open data visualizer” directly from the search page or Clustering page) and navigate to the “Communication” tab on the Data Visualizer page.
- In the resulting visualization by default, you will see the top 10 individuals (i.e., Name nodes) involved in communications about the selected topics within the underlying search. You can increase the number of name nodes displayed to identify more people that are involved in communications about the selected topics.
Learn more about the communication patterns of specific individuals
In this scenario, you have already identified one or more people of interest in the default top 10 “Names” node type visualization and want to learn more about their communication patterns. Specifically, you want to know who else these people of interest are frequently talking to during specific date ranges.
- Select nodes for one or more people of interest and click “Add to filter” in the side menu.
- Click the name node for each person of interest to see a breakdown of who they were talking to and how often. You can increase the number of neighboring name nodes displayed to identify more people that the specified name nodes were speaking to.
- Expand the date filter and select relevant date ranges to understand how communications involving specific people of interest evolved over time.
Find anomalies in communication patterns
With Communication Visualizer, you can also quickly spot when individuals are using private email addresses to send/receive emails.
- Change the node type for your visualization to Domains so you can see the top email domains which can indicate the organizations that have communicated most frequently in the underlying search.
- Select the edge between two Domain nodes of interest (e.g., @enron.com and @aol.com) and click “Add to filter” in the side menu.
- Change the node type for your visualization to Names so you can see which individuals were involved in emails exchanged between the two selected Domain nodes of interest.
- Select Name nodes of interest (i.e., persons of interest) and click “Add to filter” in the side menu to narrow your search then click “View results table” to review the relevant emails.
0 Comments