| 1. |
- Mohammadinodooshan, Alireza, 1983-
(författare)
-
Data-driven Contributions to Understanding User Engagement Dynamics on Social Media
- 2024
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Social media platforms have fundamentally transformed the way information is produced, distributed, and consumed. News digestion and dissemination are not an exception. A recent study by the Pew Research Center highlights that 53% of Twitter (renamed X) users, alongside notable percentages on Facebook (43%), Reddit (38%), and Instagram (34%), rely on these platforms for their daily news. Unfortunately, not all news is reliable and unbiased, which poses a significant societal challenge. Beyond news, content posted by influencers can also play an important role in shaping opinions and behaviors.Indeed, how users engage with different classes of content (including unreliable content) on social media can amplify their visibility and shape public perceptions and debates. Recognizing this, prior research has studied different aspects of user engagement dynamics with varying classes of content. However, several unexplored dimensions remain. To better understand these dynamics, this thesis addresses part of this research gap through eight comprehensive studies across four key dimensions, where we place particular focus on news content.The first dimension of this thesis presents a large-scale analysis of users' interactions with news publishers on Twitter. This analysis provides a fine-grained understanding of engagement patterns with various classes of publishers, with key findings indicating elevated engagement rates among unreliable news publishers. The second dimension examines the dynamics of interaction patterns between public and private (less public) sharing of news articles on Facebook. This dimension highlights deeper user engagement in private contexts compared to the public sphere, with both spheres showing the highest interaction levels with highly unreliable content. The third dimension investigates the drivers of popularity among news tweets to understand what makes some tweets more/less successful in gaining user engagement. For instance, this analysis reveals the negative impact of analytic language on user engagement, with the biggest engagement declines observed among unreliable publishers. Finally, the thesis emphasizes the importance of temporal dynamics in user engagement. For example, exploring the temporal user engagement with different news classes over time, we observe a positive correlation between the reliability of a post and the early interactions it receives on Facebook. While the thesis quantitatively assesses the effects of reliability across all dimensions, it also places additional focus on the role of bias in the observed patterns.These and other insights presented in the thesis offer actionable insights that can benefit multiple stakeholders, providing policymakers and content moderators with a comprehensive perspective for addressing the spread of problematic content. Moreover, platform designers can leverage the insights to build features that promote healthy online communities, while news outlets can use them to tailor content strategies based on target audiences, and individual users can use them to make informed decisions. Although the thesis has inherent limitations, it deepens our current understanding of engagement dynamics to foster a more secure and trustworthy social media experience that remains engaging.
|
|
| 2. |
- Andersson, Olov, 1979-
(författare)
-
Learning to Make Safe Real-Time Decisions Under Uncertainty for Autonomous Robots
- 2020
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Robots are increasingly expected to go beyond controlled environments in laboratories and factories, to act autonomously in real-world workplaces and public spaces. Autonomous robots navigating the real world have to contend with a great deal of uncertainty, which poses additional challenges. Uncertainty in the real world accrues from several sources. Some of it may originate from imperfect internal models of reality. Other uncertainty is inherent, a direct side effect of partial observability induced by sensor limitations and occlusions. Regardless of the source, the resulting decision problem is unfortunately computationally intractable under uncertainty. This poses a great challenge as the real world is also dynamic. It will not pause while the robot computes a solution. Autonomous robots navigating among people, for example in traffic, need to be able to make split-second decisions. Uncertainty is therefore often neglected in practice, with potentially catastrophic consequences when something unexpected happens. The aim of this thesis is to leverage recent advances in machine learning to compute safe real-time approximations to decision-making under uncertainty for real-world robots. We explore a range of methods, from probabilistic to deep learning, as well as different combinations with optimization-based methods from robotics, planning and control. Driven by applications in robot navigation, and grounded in experiments with real autonomous quadcopters, we address several parts of this problem. From reducing uncertainty by learning better models, to directly approximating the decision problem itself, all the while attempting to satisfy both the safety and real-time requirements of real-world autonomy.
|
|
| 3. |
- Magnusson, Måns, 1981-
(författare)
-
Scalable and Efficient Probabilistic Topic Model Inference for Textual Data
- 2018
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Probabilistic topic models have proven to be an extremely versatile class of mixed-membership models for discovering the thematic structure of text collections. There are many possible applications, covering a broad range of areas of study: technology, natural science, social science and the humanities.In this thesis, a new efficient parallel Markov Chain Monte Carlo inference algorithm is proposed for Bayesian inference in large topic models. The proposed methods scale well with the corpus size and can be used for other probabilistic topic models and other natural language processing applications. The proposed methods are fast, efficient, scalable, and will converge to the true posterior distribution.In addition, in this thesis a supervised topic model for high-dimensional text classification is also proposed, with emphasis on interpretable document prediction using the horseshoe shrinkage prior in supervised topic models.Finally, we develop a model and inference algorithm that can model agenda and framing of political speeches over time with a priori defined topics. We apply the approach to analyze the evolution of immigration discourse in the Swedish parliament by combining theory from political science and communication science with a probabilistic topic model.
|
|
| 4. |
- Sidén, Per, 1987-
(författare)
-
Scalable Bayesian spatial analysis with Gaussian Markov random fields
- 2020
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Accurate statistical analysis of spatial data is important in many applications. Failing to properly account for spatial autocorrelation may often lead to false conclusions. At the same time, the ever-increasing sizes of spatial datasets pose a great computational challenge, as many standard methods for spatial analysis are limited to a few thousand data points.In this thesis, we explore how Gaussian Markov random fields (GMRFs) can be used for scalable analysis of spatial data. GMRFs are closely connected to the commonly used Gaussian processes, but have sparsity properties that make them computationally cheap both in time and memory. The Bayesian framework enables a GMRF to be used as a spatial prior, comprising the assumption of smooth variation over space, and gives a principled way to estimate the parameters and propagate uncertainty.We develop new algorithms that enable applying GMRF priors in 3D to the brain activity inherent in functional magnetic resonance imaging (fMRI) data, with millions of observations. We show that our methods are both faster and more accurate than previous work. A method for approximating selected elements of the inverse precision matrix (i.e. the covariance matrix) is also proposed, which is important for evaluating the posterior uncertainty. In addition, we establish a link between GMRFs and deep convolutional neural networks, which have been successfully used in countless machine learning tasks for images, resulting in a deep GMRF model. Finally, we show how GMRFs can be used in real-time robotic search and rescue operations, for modeling the spatial distribution of injured persons.
|
|
| 5. |
- Rodriguez Déniz, Héctor, 1982-
(författare)
-
Bayesian Models for Spatiotemporal Data from Transportation Networks
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Urbanization has caused a historical transformation at a global scale, and humanity is moving towards a fully connected society where cities will concentrate population, infrastructure and economic activity. A key element in the cities’ infrastructure is the transportation system, as it facilitates the mobility of people and goods. Transportation systems are constantly generating data from, e.g., GPS, sensors and cameras, and the statistical modeling is challenging due to the complex structure and dynamics of the system, and the inherent uncertainty. In this thesis, we develop Bayesian models with applications to transportation. We specifically focus on models that can be trained on spatiotemporal data coming from transport networks to make predictions on, e.g., bus delays or the actual network topology. Special attention has been given to model scalability issues and uncertainty quantification. We have used real-world data from transportation systems in every study to keep a balance between statistical rigor, novelty, and applicability. The thesis consists of four papers. The first study presents a state-of-the-art probabilistic latent network model to forecast multilayer dynamic graphs. The model uses stochastic blockmodeling to reduce the computational burden, and is illustrated on a sample of 10-year data from four major airlines within the US air transportation system. In the second paper, we develop a robust model for real-time bus travel time prediction that departs from Gaussian assumptions by using Student-t errors, and show how Bayesian inference naturally allows for predictive uncertainty quantification in a highly stochastic environment. Experiments are performed using data from high-frequency buses in Stockholm, Sweden. The third paper shows the potential of multi-output Gaussian processes to tackle network-wide travel time prediction in an urban area. We develop a responsive online model based on a coregionalized covariance and test its accuracy on real data from GPS-equipped taxis. Finally, we propose a novel regularization strategy for the vector autoregressive model that is based on a graphical spike-and-slab prior, and present a case study with real airline delay data to assess its predictive performance and analyze network patterns related to the propagation of delays across airports.
|
|
| 6. |
- Svahn, Caroline
(författare)
-
Prediction Methods for High Dimensional Data with Censored Covariates
- 2022
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- While access to data steadily increases, not all data are straight-forward to use for prediction. Censored data are common in several industrial scenarios, and typically arise when there are some limitations to measuring equipment such as for instance concentration measuring equipment in chemistry or signal receivers in signal processing. In this thesis, we take several angles to censored covariate data for prediction problem. We explore the impact on both covariates and the response when the censored covariates are imputed. We consider linear approaches as well as non-linear approaches, and we explore how both frequentist models as well as Bayesian models perform with censored covariate data. While the focus is using the imputed covariate data for prediction, we also investigate model parameter inference and uncertainty inferred by the imputations. We use real, censored covariate telecommunications data for prediction with some of the most commonly used prediction models and evaluate the performance when single imputations are made. We propose a selective multiple imputation approach which is suitable for high dimensional data that perform well with heavy censoring. We take a Bayesian linear regression approach leveraging information from auxiliary variables using multivariate regression and introduce multivariate draws from conditional distributions to update censored values in the covariates. We fnally offer a bridge between the fexibility of Neural Networks and the probabilistic nature of Bayesian methods by taking a Variational Autoencoder approach and introducing Zero-Infated Truncated Gaussian likelihoods for the covariates to better ft the censored distributions.
|
|
