Modeling the Online Spread of Ideas in a Finite Attention Environment

Calderon, Pio Gabrielle Battad

Modeling the Online Spread of Ideas in a Finite Attention Environment

Calderon, Pio Gabrielle Battad

Permalink

Publication Type:: Thesis
Issue Date:: 2024

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (8.67 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Calderon, Pio Gabrielle Battad
dc.date.accessioned	2025-09-15T02:49:24Z
dc.date.available	2025-09-15T02:49:24Z
dc.date.issued	2024
dc.identifier.uri	http://hdl.handle.net/10453/189837
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	Online social systems are challenging to model due to the heterogeneity of human behavior influenced by diverse cultural, economic, historical, and political factors. Additionally, measurements from these systems tend to be incomplete and noisy due to platform constraints and unpredictable human actions. Despite these challenges, online social systems are governed by foundational mechanisms that can be modeled to gain insights into collective behaviors. This thesis explores models of the spread of ideas in online social systems with three primary objectives: learn the latent mechanisms that can explain the observed noisy data, predict future online diffusions, and evaluate the impact of external interventions. The first contribution is the Opinion Market Model (OMM), a two-tier system of the online opinion ecosystem that jointly captures inter-opinion interactions and the impact of positive interventions in a finite attention environment. The OMM outperforms state-of-the-art models in understanding opinion dynamics and can be leveraged as a testbed to evaluate media as an intervention to redirect attention from extremist to moderate opinions. The second contribution is the Bayesian Mixture Hawkes (BMH) model, a hierarchical mixture model of separable Hawkes process that can jointly capture the influence of source, content, and cascade-level factors on the spread dynamics of online items. The BMH model excels in predicting content popularity in the cold-start setup and can differentiate the impact of different headline styles across publishers. The third contribution is the development of the Partially Censored Multivariate Hawkes Process (PCMHP), which addresses the challenge of fitting the self- and cross-exciting multivariate Hawkes process in the partially interval-censored setting. The PCMHP can model cross-platform data with limited availability, such as mixed event-timestamp data and daily-aggregated counts, and outperforms existing models in predicting YouTube popularity. This thesis advances our understanding of the spread of ideas in online social systems, providing robust models for explaining, predicting and influencing online behavior.	en_US.UTF-8
dc.format	Thesis (PhD)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/189837/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2024 Pio Gabrielle Battad Calderon
dc.rights	au.edu.uts.lib/cph
dc.title	Modeling the Online Spread of Ideas in a Finite Attention Environment	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

Online social systems are challenging to model due to the heterogeneity of human behavior influenced by diverse cultural, economic, historical, and political factors. Additionally, measurements from these systems tend to be incomplete and noisy due to platform constraints and unpredictable human actions. Despite these challenges, online social systems are governed by foundational mechanisms that can be modeled to gain insights into collective behaviors. This thesis explores models of the spread of ideas in online social systems with three primary objectives: learn the latent mechanisms that can explain the observed noisy data, predict future online diffusions, and evaluate the impact of external interventions. The first contribution is the Opinion Market Model (OMM), a two-tier system of the online opinion ecosystem that jointly captures inter-opinion interactions and the impact of positive interventions in a finite attention environment. The OMM outperforms state-of-the-art models in understanding opinion dynamics and can be leveraged as a testbed to evaluate media as an intervention to redirect attention from extremist to moderate opinions. The second contribution is the Bayesian Mixture Hawkes (BMH) model, a hierarchical mixture model of separable Hawkes process that can jointly capture the influence of source, content, and cascade-level factors on the spread dynamics of online items. The BMH model excels in predicting content popularity in the cold-start setup and can differentiate the impact of different headline styles across publishers. The third contribution is the development of the Partially Censored Multivariate Hawkes Process (PCMHP), which addresses the challenge of fitting the self- and cross-exciting multivariate Hawkes process in the partially interval-censored setting. The PCMHP can model cross-platform data with limited availability, such as mixed event-timestamp data and daily-aggregated counts, and outperforms existing models in predicting YouTube popularity. This thesis advances our understanding of the spread of ideas in online social systems, providing robust models for explaining, predicting and influencing online behavior.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/189837