Building Social Media Observatories for Monitoring Online Opinion Dynamics

Abstract

Social media house a trove of relevant information for the study of online opinion dynamics. However, harvesting and analyzing the sheer overload of data that is produced by these media poses immense challenges to journalists, researchers, activists, policy makers, and concerned citizens. To mitigate this situation, this article discusses the creation of (social) media observatories: platforms that enable users to capture the complexities of social behavior, in particular the alignment and misalignment of opinions, through computational analyses of digital media data. The article positions the concept of “observatories” for social media monitoring among ongoing methodological developments in the computational social sciences and humanities and proceeds to discuss the technological innovations and design choices behind social media observatories currently under development for the study of opinions related to cultural and societal issues in European spaces. Notable attention is devoted to the construction of Penelope: an open, web-services-based infrastructure that allows different user groups to consult and contribute digital tools and observatories that suit their analytical needs. The potential and the limitations of this approach are discussed on the basis of a climate change opinion observatory that implements text analysis tools to study opinion dynamics concerning themes such as global warming. Throughout, the article explicitly acknowledges and addresses potential risks of the machine-guided and human-incentivized study of opinion dynamics. Concluding remarks are devoted to a synthesis of the ethical and epistemological implications of the exercise of positioning observatories in contemporary information spaces and to an examination of future pathways for the development of social media observatories.

Opinion dynamics in social networks: From models to data

“…Models of opinion dynamics help clarify the interplay between human behavior and artificial intelligence, hinting at how algorithmic bias can promote echo chambers and opinion polarization online (Santos et al, 2021), or instead help disseminate information across structural divides, depending on the mechanisms of interaction assumed by the models (Sîrbu et al, 2019;Perra and Rocha, 2019;Peralta et al, 2021b,a). Online social networks are expected to provide empirical background for studying opinion dynamics by means of natural language processing (Willaert et al, 2020) Modeling opinion dynamics in social networks shows the potential of an interdisciplinary approach to understand the emergence of collective behaviour from the actions and interactions of individuals. How, despite the multifaceted complexity of people's psychology and the many sociopolitical contexts humans engage in, we can still link individual preferences to global states of consensus and polarization of opinions.…”

Classification of Negative Information on Socially Significant Topics in Mass Media

“…For example, in a task comparing Turkish and English, the corpora of news related to science is considered. Moreover, media monitoring includes such tasks as social behavior, public opinion identification [11,12], and online sales trend analysis [13] (task 2), and comparison of preferences and characteristics of population segments (task 3). For example, Macharia [14] analyses gender inequality.…”

An Opinion Facilitator for Online News Media

“…This interdisciplinary project explores a range of text mining techniques and advances in computational methods for language analysis to map opinion landscapes on cases such as migration, antisemitism, antagonistic right-wing discourse, and climate change. These methods, resulting from the project's fundamental research, are made openly available in an ecosystem of tools and techniques for computational social science called Penelope (Penelope, 2019a;Willaert et al, 2020). The Penelope ecosystem thus comprises modules (individual data-analytical components, often without a user interface) and observatories, which chain together components and have a user interface that allows for the exploration of a specific theme, such as language propagation on social media (Willaert et al, 2021).…”

“…Applied to a dataset of English newspaper articles on climate change similar to the one that will be central to the present paper, the computational construction grammar approach achieves a word-level F1 score of 78.5%, outperforming a commonly used approach based on conditional random fields (CRFs). As this approach offers a stable method for extracting instances of semantic frames, in particular instances of the causation frame, it has previously been used to analyze belief systems expressed in online news environments, such as the argumentative domains surrounding energy technologies expressed in comments on the news website of The Guardian (Willaert et al, 2020;Willaert et al, Under review) 1 .…”