[Paper Review] Entropy in Social Networks

John L. Pfaltz|arXiv (Cornell University)|Dec 12, 2012

Complex Network Analysis Techniques27 references3 citations

TL;DR

This paper introduces a non-statistical, closure-based mathematical framework to model entropy in dynamic social networks, using closure operators and continuous transformations. It shows that continuous changes preserve separation and connectivity, while creating new links—especially triadic closures—is discontinuous and energy-intensive, leading to a breakdown of complex structures into triadically closed clusters, mirroring sociological observations of network evolution.

ABSTRACT

We introduce the concepts of closed sets and closure operators as mathematical tools for the study of social networks. Dynamic networks are represented by transformations. It is shown that under continuous change/transformation, all networks tend to "break down" and become less complex. It is a kind of entropy. The product of this theoretical decomposition is an abundance of triadically closed clusters which sociologists have observed in practice. This gives credence to the relevance of this kind of mathematical analysis in the sociological context.

Motivation & Objective

• To develop a non-statistical mathematical model of network entropy based on closure operators and transformations.
• To investigate how continuous transformations affect network structure and connectivity in dynamic social networks.
• To explore the relationship between network complexity, closure, and the emergence of triadic clusters.
• To test whether closure-based continuity and discontinuity can explain observed network evolution patterns, such as triadic closure.

Proposed method

• Uses closure operators φ defined by three axioms: extensive (Y ⊆ Y.φ), monotone (Y ⊆ Z ⇒ Y.φ ⊆ Z.φ), and idempotent (Y.φ.φ = Y.φ).
• Models social processes as monotone and continuous transformations f between closure systems (P, φ) → (P′, φ′), with continuity defined by Y.φ.f ⊆ Y.f.φ′.
• Applies the concept of subsumption: if y ∈ Y.φ, then y is subsumed by the closure, and such points can be removed iteratively to find irreducible subgraphs.
• Identifies chordless k-cycles (k ≥ 4) as structural building blocks of complex networks, with chordal graphs (no such cycles) being simpler.
• Uses Proposition 11 to show that creating new links between separated sets (e.g., adding edge (h,i)) is discontinuous, requiring energy.
• Compares results to Granovetter’s theory of strong ties and bridges, showing that triadic closure is discontinuous, while edge removal is typically continuous.

Experimental results

Research questions

• RQ1 How do continuous transformations affect closure relations and connectivity in dynamic social networks?
• RQ2 What mathematical conditions define the creation of new links between separated network components?
• RQ3 Why do triadic closures emerge in social networks, and is this process continuous or discontinuous?
• RQ4 Can closure operators and transformation continuity explain the observed breakdown of complex network structures into simpler, clustered forms?

Key findings

• Continuous transformations preserve separation and connectivity, meaning that existing clusters and connections are maintained under smooth change.
• Creating new links between separated components—such as adding edge (h,i) to connect clusters—is a discontinuous process, as shown by Proposition 11.
• Triadic closure, a well-known sociological phenomenon, is inherently discontinuous and requires external energy, aligning with the concept of entropy in network evolution.
• The emergence of chordless k-cycles (k ≥ 4) is a key indicator of complex network structure, and their presence or absence defines whether a network is chordal or not.
• Networks undergoing continuous change tend to decompose into triadically closed clusters, suggesting a natural tendency toward entropy-driven simplification.
• The model provides a discrete, non-statistical alternative to traditional entropy measures, grounding network complexity in closure and transformation theory.