Probabilistic neighbors' selection competes with confirmation bias in a bounded confidence model

Chiara Giaquinta; Laura Hernández; David Chavalarias

Home/Archive/arXiv

arXiv · 2610.07971

Probabilistic neighbors' selection competes with confirmation bias in a bounded confidence model

Chiara Giaquinta·Laura Hernández·David Chavalarias

View original source ↗Read source document ↗Export metadata ↓

Abstract

In this work, we investigate three modified versions of the classic Hegselmann-Krause opinion dynamics model, incorporating features that are typical of many real-world systems, such as uncertainty in the selection of interacting agents and a weighted evaluation of the relevance of their opinions in the influence function to enhance confirmation bias. Through extensive simulations across different network topologies, ranging from stylized network models (Barabási-Albert, Erdős-Rényi, and Watts-Strogatz networks) to empirical networks with community structure, we identify the influence of each modification on opinion evolution and convergence. Our findings reveal that they exert opposite effects on the bounded confidence threshold required for consensus. We further explore an extension based on a data-driven opinion initialization on the empirical networks, where initial opinions are drawn from Gaussian distributions specific to each detected community. While the qualitative effects of the three modified models remain consistent, this new initialization strategy reveals distinct dynamics within the network communities. These insights provide a new and comprehensive perspective on how realistic variations of the Hegselmann-Krause model, in terms of both interaction rules and initial opinion distributions, affect opinion dynamics and shed light on the mechanisms of consensus formation within structured communities.

Explore related subjects

physics.soc-ph

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Chiara Giaquinta, Laura Hernández, David Chavalarias. 2026-10-06. Probabilistic neighbors' selection competes with confirmation bias in a bounded confidence model. https://arxiv.org/abs/2610.07971

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

The Hidden Geometry of Global Aid: How Money Moves Through 10 Million Transactions

International aid is usually described as flows between countries, hiding the organisations money passes through. We reconstruct global aid's organisational network from 10.9 million International Aid Transparency Initiative records, tracing over one million payments worth about US\$1 trillion between 22,773 organisations. First, centrality, measured as how far a funder's money can travel downstream, is not size. Relative to their number of partners, giant bilateral donors (USAID, UK FCDO) are only as central as expected, while the most central positions belong to foundations (Gates, Hewlett), research organisations (ODI, University of Oxford) and delivery contractors (Mott MacDonald). Second, principal component analysis of eleven flow measures separates organisations along three lines: whether they mainly give or receive, how much money flows in, and how reliant they are on few partners. Humanitarian organisations emerge from the flows alone, tending to handle larger inflows. Third, almost all organisations form one connected network, yet money moves down long, one-way chains. Among five comparison networks, aid most resembles international banking, with a few large hubs feeding many small players, but has far fewer loops and is more strictly top-down. This suggests a system with little redundancy, where losing a major funder or intermediary could cut off organisations downstream, a pressing question after recent donor cuts.

physics.soc-ph↗

arXiv2026

Beyond expressiveness in pairwise and higher-order models

The debate over pairwise and higher-order models is often framed as a contest of expressive power. This framing is misleading. A graph with arbitrary multivariate node functions can reproduce the node-level dynamics of many hypergraph models, but the hypergraph's grouping information does not thereby disappear: it may simply move from the structure to the rule. We separate four notions frequently conflated here: structural representation, functional representability, statistical identifiability, and mechanistic adequacy. We then prove that interaction order, the number of variables that must act jointly in some term of any additive decomposition of a node's update, is the same in every exact representation: no change of structural language can lower it. As a description-length problem, at the unrestricted algorithmic level a fixed compiler redistributes information between structure and rule at constant overhead, so expressiveness alone cannot privilege either language. Preferences arise only relative to explicit model classes, code families, and data. This yields an operational minimum-description-length criterion combining structural cost, conditional rule cost, and imperfect fit. Within it, for $M$ disjoint groups of size $k$, the clique projection's edge list is asymptotically $k-1$ times longer than the hyperedge list it replaces: the projection encodes the same grouping at higher cost. Examples from diffusion, Boolean dynamics, ecology, and ambiguous projections give graph-preferred, hypergraph-preferred, and unresolved cases; for bipartite and multilayer lifts, cost, fit, and identifiability tie, and the choice turns on which entities are posited as primitive. The position is symmetric: higher-order structure should not be inferred from phenomenology alone, nor does a graph's ability to emulate a system make it the most parsimonious or adequate description.

physics.soc-ph↗

arXiv2026

Who Waits the Longest? Unequal Paths to Legal Stability among Migrants and Refugees

Legal systems shape not only the recognition of migrants and refugees but also the pace and stability of their integration. Refugees often shift between multiple legal classifications, a process we refer to as the "legal journey". This journey is frequently prolonged and uncertain. Using a network-based approach, we analyze legal transitions for over 350,000 migrants in Austria (2022-2024). EU nationals such as Germans reach stable residence in under four months; refugees from conflict-affected regions face far longer and more uncertain journeys, ranging from two months for Ukrainians to nine months for Syrians and 20 months for Afghans. Refugee women, especially from Syria and Afghanistan, are more likely to gain protection. Afghan women reach stability in 14 months on average, less than half the 30 months Afghan men wait. We also find that those who cross the border without going through official border controls face higher exit rates and lower chances of securing a stable status. We show that legal integration is not a uniform process, but one structured by institutional design, entry mode, and unequal timelines.

physics.soc-ph↗