We have an evolving workflow for discovering hidden structure, such as patterns, in networks from various sources. These are preprocessed into "aspects" highly localized perspectives that are then browsed and reassembled into meaningful structures.
See Discovery Workflow outline.
# Access
We fetch and interpret wiki page json for many reasons. Here we collect and distribute as an ES6 module our functions found most useful. github 
index.js
We sometimes handcode graph's nodes and relations in an abbreviated notion called "freeform".
We have the opportunity to carefully read "Moldable Development Patterns" by Nierstrasz and Gîrba. We encoded this in free text in Code items interpreted with an html script. script github
We have understood a sprawling service coded in ruby and shell script by annotating important operations with an easily recognizable comment encoding relevant nodes and their relationship in the subsequent code.
I have added automatically maintained dataflow diagrams to the federation search. I think the approach I have used has broad application for anyone who maintains complicated software with many interacting parts.
# Features
We may find aspects in natural structure as we have with the zones and regions of radio network routing codes.
Routing is guided by numbered zones, each divided into numbered regions which themselves group a number of stations. github
Recent activity can be usefully divided into "temporal" aspects of days, weeks or moths.
Mech explores a wiki neighborhood. The WALK block travels in increments called aspects.
Distinct properties of network nodes can be rank-ordered by popularity to make top-ten lists. Members and their immediate neighbors make good aspects.
We discover stories within the noosphere by collecting recent components of the markov blanket for likely hubs. github