Toward a Standard Approach for Echo Chamber Detection: Reddit Case Study

Morini, Virginia; Pollacci, Laura; Rossetti, Giulio

doi:10.3390/app11125390

Cited by 27 publications

(14 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By selecting the right tools, such as multilingual language models, transliteration, or machine translation, the same approach can be used to quantify the content similarity of any two communities of any two social networks. of the figure shows poll results in Croatia [2] , Serbia [3] , Republika Srpska (part of Bosnia and Herzegovina governed by Bosnian Serbs) [4] , and Montenegro [5] on the following question: "Who do you support in the Ruso-Ukrainian conflict?". The bottom part shows a map of Eastern Europe, with color marks on the BCMS-speaking countries covered in this research, Ukraine (yellow) and Russia (red).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Content-based Comparison of Communities in Social Networks: Ex-Yugoslavian Reactions to the Russian Invasion of Ukraine

Evkoski

Novak

Ljubešić

2023

Preprint

View full text Add to dashboard Cite

We propose an approach for comparing communities in social networks based on used hashtags, shared news sources, and discussed topics and their sentiment, both derived through transformer-based language modeling. We apply this multifaceted comparison to study the reactions of the Ex-Yugoslavian retweet communities to the Russian invasion of Ukraine. Our results show that even though Ex-Yugoslavian countries share a common macro-language, their retweet communities respond very heterogeneously to the invasion. We demonstrate the limitations of observing communities using topology only and illustrate the benefits and complementarity of including content when analyzing social networks. As an example, we have identified a community linked to the Serbian governing power that deliberately avoids the invasion topic, which would not be possible by analyzing hashtags only. We conclude that a multifaceted approach provides a more nuanced understanding of social network communities that can be applied to gain insights into complex social events beyond the case study presented in this paper.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Most often, the resulting communities are purely topological [4], including only the interactions between nodes into account. Yet, their interpretation is typically qualitative in nature, relying on additional contextual data that is not, or cannot be, represented in the network's topology [5,6,7,8].…”

Section: Introductionmentioning

confidence: 99%

Content-based Comparison of Communities in Social Networks: Ex-Yugoslavian Reactions to the Russian Invasion of Ukraine

Evkoski

Novak

Ljubešić

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…As future works, we plan to apply -Conformity for hot topics in computational social science, as in multi-scale echo-chamber identification [22], or for monitoring users in sensible online communities, e.g., mental health-related ones [12]. We expect that the same mechanisms applied on Bitcoin network could allow us to identify cliques of actors working together to manipulate the market or the formation of national markets, for instance in a country adopting Bitcoin as local tender such as El Salvador.…”

Section: Discussionmentioning

confidence: 99%

$$\Delta $$-Conformity: multi-scale node assortativity in feature-rich stream graphs

Citraro

Milli

Cazabet

et al. 2022

Int J Data Sci Anal

Self Cite

View full text Add to dashboard Cite

Multi-scale strategies to estimate mixing patterns are meant to capture heterogeneous behaviors among node homophily, but they ignore an important addendum often available in real-world networks: the time when edges are present and the time-varying paths that edges form accordingly. In this work, we go beyond the assumption of a static network topology to propose a multi-scale, path- and time-aware node homophily estimator specifically tied for feature-rich stream graphs: $$\Delta $$ Δ -Conformity. Our measure can capture the homogeneous/heterogeneous tendency of nodes’ connectivity along a period of time $$\Delta $$ Δ starting from a given moment in time. Results on face-to-face interaction networks suggest it is possible to track changes in social mixing behaviors that coincide with contextually reasonable everyday patterns, e.g., medical staff disassortative behavior when exposed to patients. In a different domain, that of the Bitcoin Transaction Network, we capture relationships between the quantity of money sent from (and to) different categories/continents and their respective mixing trends over time. All these insights help us to introduce $$\Delta $$ Δ -Conformity as a suitable solution for understanding temporal homophily by capturing the mixing tendency of entities embedded in fine-grained evolving contexts.

show abstract

“…These recommendations connect users with needed items that they would have had to spend time searching for [50]. Although these recommendations may lead to harmful effects such as echo chambers [34,51], techniques to identify and mitigate harms are improving [38,64].…”

Section: Conclusion Limitations and Broader Impactmentioning

confidence: 99%

On the Generalizability and Predictability of Recommender Systems

McElfresh¹,

Khandagale²,

Valverde³

et al. 2022

Preprint

View full text Add to dashboard Cite

While other areas of machine learning have seen more and more automation, designing a high-performing recommender system still requires a high level of human effort. Furthermore, recent work has shown that modern recommender system algorithms do not always improve over well-tuned baselines. A natural follow-up question is, "how do we choose the right algorithm for a new dataset and performance metric?" In this work, we start by giving the first large-scale study of recommender system approaches by comparing 18 algorithms and 100 sets of hyperparameters across 85 datasets and 315 metrics. We find that the best algorithms and hyperparameters are highly dependent on the dataset and performance metric, however, there are also strong correlations between the performance of each algorithm and various meta-features of the datasets. Motivated by these findings, we create RecZilla, a meta-learning approach to recommender systems that uses a model to predict the best algorithm and hyperparameters for new, unseen datasets. By using far more meta-training data than prior work, RecZilla is able to substantially reduce the level of human involvement when faced with a new recommender system application. We not only release our code and pretrained RecZilla models, but also all of our raw experimental results, so that practitioners can train a RecZilla model for their desired performance metric: https://github.com/naszilla/reczilla.

show abstract

Toward a Standard Approach for Echo Chamber Detection: Reddit Case Study

Cited by 27 publications

References 41 publications

Content-based Comparison of Communities in Social Networks: Ex-Yugoslavian Reactions to the Russian Invasion of Ukraine

Content-based Comparison of Communities in Social Networks: Ex-Yugoslavian Reactions to the Russian Invasion of Ukraine

$$\Delta $$-Conformity: multi-scale node assortativity in feature-rich stream graphs

On the Generalizability and Predictability of Recommender Systems

Contact Info

Product

Resources

About