Situating Visual Network Analysis (PhD thesis)

Situating Visual Network Analysis is my PhD thesis, just accepted for defense, scheduled for June 1 at 12:00 CET in Copenhagen. I will provide additional details when I have them.

In the meanwhile here is the manuscript for download if you want to take a look:

Download the manuscript (PDF)

Abstract

Visual network analysis (VNA) is the practice of analyzing networks by visual means. In this dissertation, I account for this practice and the techniques involved by focusing on force-directed node placement algorithms, the most popular strategy for drawing network maps. I explore the question of what we see when we look at networks, address some of the criticism faced by network visualization, and reflect on the role of the layout algorithm in the visual mediation of the network’s topological structure. My argument unfolds in six theses: (1) VNA consists of practices that are only partially determined by the graph-drawing and data-visualization literature; (2) some visualizations, including network maps, prompt a visual inquiry into the meaning of emergent patterns as contributing to their apparent self-evidence; (3) for historical reasons, the graph-drawing literature mainly promotes an interpretation regime adapted to small networks (diagrammatic), while practices partially shifted in the 2000s to large networks (topological interpretation regime); (4) some issues with reading network maps can be attributed to the misalignment between our visual cognition and the computational standpoint, notably the notion of a group; (5) the existing justifications of algorithm designers do not provide a compelling explanation of what we see in networks; and (6) the literature on community detection focuses on clear-cut clusters, while force-driven placement algorithms make visible other non-clear-cut community structures.


One thought on “Situating Visual Network Analysis (PhD thesis)”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.