A interactive visualization of the references between diplomatic cables.
Please use a recent Chrome or Safari. Firefox does display it correctly, but is missing the "jump" functionality.
The cables.csv contains a field for referencing related cables via MRN. In the cables this is often called "REFTEL" and the following notation is used in the cable body.
REF: TEHRAN 1091: TEHRAN 263: MOSCOW 1603
REF: A) HARARE 17 B) HARARE 394 C) 87 HARARE 6985 D) HARARE 240
Around ~232300 of these references exist and they form a large network (or so called graph).
Not all cables belong to a single large network but rather form smaller groups of networks. These smaller groups are also known as clusters or connected components.
Some of these clusters form a star which can be a indication for a type of "action request" that was sent to multiple embassies.
The largest cluster (giant) contains ~100000 cables with ~143000 references. A graph method called "community detection" can be used to split the giant cluster into smaller communities.
For every community, cluster or star a interactive visualization is generated.
From the total of 251287 cables ~56000 additional cables are referenced but not included in the cables.csv. These missing cables are also shown in the graph.
For some of the missing cables it is possible to estimate the date of creation. This estimated date is included.
The visualization contains some indicators of graph and cable properties.
cable2graph source code (GPLv3).
cable2graph comes with tools and data useful for research on the cables.csv
@datapornstar can haz moar data pr0n?