Community Interaction and Conflict on the Web

By Srijan Kumar, William L. Hamilton, Jure Leskovec, Dan Jurafsky. The Web Conference, 2018

402 citations recorded by the source corpus.

Users organize themselves into communities on web platforms. These communities can interact with one another, often leading to conflicts and toxic interactions. However, little is known about the mechanisms of interactions between communities and how they impact users. Here we study intercommunity interactions across 36,000 communities on Reddit, examining cases where users of one community are mobilized by negative sentiment to comment in another community. We show that such conflicts tend to be initiated by a handful of communities---less than 1% of communities start 74% of conflicts. While conflicts tend to be initiated by highly active community members, they are carried out by significantly less active members. We find that conflicts are marked by formation of echo chambers, where users primarily talk to other users from their own community. In the long-term, conflicts have adverse effects and reduce the overall activity of users in the targeted communities. Our analysis of user interactions also suggests strategies for mitigating the negative impact of conflicts---such as increasing direct engagement between attackers and defenders. Further, we accurately predict whether a conflict will occur by creating a novel LSTM model that combines graph embeddings, user, community, and text features. This model can be used to create an early-warning system for community moderators to prevent conflicts. Altogether, this work presents a data-driven view of community interactions and conflict, and paves the way towards healthier online communities.

Subfields: Artificial Intelligence, Communication, Information Systems

View Paper: https://www.semanticscholar.org/paper/9935c8f8a37360196d92221ec092ae1880d272b8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Communication

In Information Systems

Citations in the corpus

Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks (2019, background): “We use four datasets from Reddit, Wikipedia, LastFM, and a MOOC course activity for our experiments.”
Developing an online hate classifier for multiple social media platforms (2020, background): “Often, these studies involve a predictive machine learning aspect for the detection and classification of toxicity in specific communities and social media platforms [2, 5, 16, 47, 48].”
GNNExplainer: Generating Explanations for Graph Neural Networks (2019, methodology): “Conversely, we observe that discussion patterns commonly exhibit tree-like patterns (2nd row in Figure 4A), since a thread on Reddit is usually a reaction to a single topic [24].”
Relational Representation Learning for Dynamic (Knowledge) Graphs: A Survey (2019, background): “Reddit HyperLink Network 19 : Subreddit hyperlink network [127] is a directed network extracted from the posts that create hyperlinks from one subreddit to another.”
Representation Learning for Dynamic Graphs: A Survey (2019, methodology): “In particular, a model developed for CTDGs may be used for DTDGs, but the reverse is not necessarily true.”
The Pushshift Reddit Dataset (2020, background): “2019; Tan 2018), conflict in online groups (Datta and Adar 2019; Datta, Phelan, and Adar 2017; Kumar et al. 2018), the spread of technological innovations (Glenski, Saldanha, and Volkova 2019), modeling collaboration (Kasper et al.”