This limited simulation of addressees connects with the Dossier Space experiment as published by the UK Society for Computers and Law.

How to Use the Communication Network Model
This interactive model is a visual aid for exploring communication relationships in the selected Enron email material. It helps identify who appears connected to whom, how often exchanges occurred, and where a participant may sit within the wider communication network. It is an investigative aid, not a finding of fact by itself.

Important: A visible connection means that the underlying dataset contains an observed communication relationship. It does not, by itself, establish friendship, authority, knowledge, wrongdoing, or any other substantive relationship. The source messages must be examined before drawing conclusions.
Getting started

1. Look at the whole network
   Start with All displayed participants. This gives a general view of the strongest communication relationships in the reduced display.
2. Choose a participant
   Use Focus participant to select a person, for example John Dudmesh. The model will emphasise that participant and their displayed relationships.
3. Adjust the threshold
   Move Minimum exchanges upward to remove weaker relationships, or downward to reveal single and low-frequency exchanges.
4. Examine direction
   Use Direction to show all relationships, messages sent by the selected participant, or messages received by them.
   What the controls do
   Minimum exchanges
   This slider filters relationships by the number of observed exchanges. A value of 1 shows even a single observed exchange. Increasing the value progressively removes weaker relationships and makes the stronger communication structure easier to see.

Focus participant
Selecting a participant highlights that person and the relevant displayed relationships. This is the most useful control when testing a specific investigative question, such as whether a person appears peripheral or embedded within a wider communication group.

Direction
With a participant selected, All relationships shows both incoming and outgoing connections. Sent by focused participant shows observed outgoing relationships, while Received by focused participant shows incoming relationships.

Reset layout
This clears the selected participant, restores the default exchange threshold and redraws the network.

Using the network directly
Move the pointer over a node to see the participant name and basic sent/received counts. Click a node to select it. Nodes can also be dragged to make crowded areas easier to inspect. When a participant is selected, the detail area beneath the network lists that participant’s strongest displayed relationships together with exchange counts and observed date ranges.

Interpreting John Dudmesh
In the current proof-of-concept dataset, John Dudmesh is deliberately retained in the visualisation even though the model records only a very small number of observed relationships involving him. That is useful because investigative significance is not the same thing as communication frequency. A person who appears only once may be peripheral, but a single message can also provide the starting point for a wider enquiry.

To inspect John, select Dudmesh, John from the participant list and lower Minimum exchanges to 1. The visible link can then be treated as a lead to the underlying message and the surrounding documentary context.

What the network can help you ask
Does this person appear central, peripheral, or isolated in the selected material?
Who communicates most frequently with them?
Are the relationships predominantly incoming or outgoing?
Over what period do the observed exchanges occur?
Does a single apparently important message sit inside a wider communication pattern?
Does the network suggest another person, date range, project, or mailbox worth examining?
What it cannot tell you
The model does not determine motive, responsibility, seniority, knowledge or wrongdoing. It also does not establish that the available corpus is complete. Missing communications may have occurred by telephone, meeting, another email system, an unavailable mailbox, or through material not preserved in the dataset.

Network position can also be misleading. A heavily copied recipient may appear highly connected without being substantively involved, while a specialist who communicates through a narrow channel may appear unimportant despite performing a significant role.

Working rule: Use the network to generate and test investigative questions. Use the underlying evidence to support findings.
A simple investigative workflow
Start with the person or issue being investigated.
Use the network to identify potentially relevant relationships.
Reduce or increase the exchange threshold to test whether those relationships are persistent or isolated.
Check communication direction and date range.
Return to the source messages and examine their actual content and context.
Record what is directly observed separately from what is inferred.
Decide whether the person or connection warrants further investigation, should remain noted for later, or can reasonably be parked.
This page explains the proof-of-concept communication network visualisation. It should be read as an interface guide, not as an evidential finding.