Topology-derived methods for the analysis of collective trust dynamics



CTRUST aims to understand how collective trust rises and disappears in social groups. We want to identify properties of social networks that promote or block the emergence, maintenance, and decay of trust. Based on this, we want to recommend actions to group leaders, managers, and industry for the successful management of social groups by creating an online platform that disseminates and converts scientific findings to applicable guidelines.


Trust is a force that holds society together. Trust between two individuals has been extensively studied. The emergence of trust in a social group is a collective phenomenon that needs to be better understood. More than 2 billion active users interact online and leave data, allowing us to study trust in social groups.


CTRUST is based on the idea that the combined knowledge about person-to-person relations and the structure of social interactions are essential for predicting and understanding the emergence of collective trust. We will develop a new set of methods that will allow us to analyze large data sets and uncover how the structure of social interactions influences the dynamics of collective trust.


We will quantify the influence of the structure of social networks on trust dynamics by applying a unique approach based on empirical analysis, modeling, and experimentation. We will collect the data from online social groups and extract information about interactions and the exchange of emotions. Using complex network theory and novel topological methods, we will investigate how the structure affects trust dynamics and derive the essential network properties related to trust. To generalize our findings, we will develop and investigate agent-based models of trust dynamics and design and conduct the experiment.

Expected results

Key ingredients of trust are encoded in the structure of evolving social networks. We will identify them and test how well they describe and predict collective trust dynamics in a social group.


Knowledge about the factors that influence the emergence and persistence of trust in social groups is one of the central societal challenges. We will create comprehensive guidelines on managing trust in social groups for industry and the public and disseminate them to a broad audience through the project website. It is expected that our findings will profoundly impact many sectors of society.


This research is supported by the Science Fund of the Republic of Serbia, Grant No 7416, Topology-derived methods for the analysis of collective trust dynamics - CTRUST.


Project Website

Project website to keep the general public updated on the project's progress and outcomes


Integrated dataset and data descriptor paper with user activity data from several online communities such as Reddit, Stack Exchange, Twitter

Paper: Emergence of Collective Trust

Paper about the emergence of collective trust in online social groups

Data Analysis code with DIBR model

Open-source analysis code includign DIBR model implementation available on GitHub

Paper: ABM simulation

Paper on the development of ABM model of collective trust and tests of different scenarios of trust emergence

Experimental Study Protocol

Online repository containing containing the experimental protocol, software setup and the resulting data for the purposes of future replication

Paper: Experimental Study on Collective Trust

Paper about the experimental study, focusing on the differences between treatment and individual effects

Paper: Spilover Effect

Spillover effect paper that combines the results from empirical analysis, agent-based modeling, and experiment



The CTRUST team comprises researchers from three national institutions: the Institute of Physics Belgrade, the Faculty of Philosophy at the University of Novi Sad (FFUNS), and the Vinča Institute of Nuclear Sciences at the University of Belgrade (VINS).






April 2024

Data Collection Protocol for the Multi-platform Aggregated Dataset of Online Communities (MADOC)

This protocol specifies the data collection process for the Multi-platform Aggregated Dataset of Online Communities (MADOC). The resulting dataset will study different aspects of online social media dynamics, user interactions, and content across different platforms: Twitter, Bluesky, Koo, Reddit, and Voat. Link:



June 2024

Core-periphery Structure and the Dynamics of Trust in Online Communities: A Comparative Study of StackExchange, Reddit, Voat, and Koo

Conference Abstract, SUNBELT 2024. Link:



Something went wrong. Please try again.
Your message was sent, thank you!


Phone: +381 (0)11 37 13 000



Pregrevica 118
11080 Belgrade