[Paper Review] Mapping Moral Valence of Tweets Following the Killing of George Floyd

J. Hunter Priniski, Negar Mokhberian+5 more|arXiv (Cornell University)|Apr 19, 2021

Social Media and Politics13 references4 citations

TL;DR

This study analyzes over 40,000 geo-located tweets from Los Angeles following George Floyd's killing to map moral valence in Black Lives Matter discourse on Twitter. Using NLP and sentiment analysis, it identifies moral dimensions in protest-related content and links moral framing to real-world protest activity, revealing that moral valence significantly predicts protest mobilization and varies by geographic and demographic context.

ABSTRACT

The viral video documenting the killing of George Floyd by Minneapolis police officer Derek Chauvin inspired nation-wide protests that brought national attention to widespread racial injustice and biased policing practices towards black communities in the United States. The use of social media by the Black Lives Matter movement was a primary route for activists to promote the cause and organize over 1,400 protests across the country. Recent research argues that moral discussions on social media are a catalyst for social change. This study sought to shed light on the moral dynamics shaping Black Lives Matter Twitter discussions by analyzing over 40,000 Tweets geo-located to Los Angeles. The goal of this study is to (1) develop computational techniques for mapping the structure of moral discourse on Twitter and (2) understand the connections between social media activism and protest.

Motivation & Objective

• To develop computational methods for mapping moral discourse structure on Twitter.
• To examine how moral valence in social media content relates to real-world protest mobilization.
• To understand the role of moral framing in shaping national discourse around racial injustice and police violence.
• To investigate geographic and demographic variations in moral valence across protest-related tweets in Los Angeles.

Proposed method

• Collected and geo-located over 40,000 tweets related to George Floyd’s killing and the subsequent protests in Los Angeles.
• Applied NLP techniques to classify moral foundations (e.g., fairness, harm, loyalty) in tweet content using pre-trained models.
• Calculated moral valence scores by aggregating moral foundation weights to quantify the moral tone of each tweet.
• Mapped spatial and temporal patterns of moral valence across LA neighborhoods to identify high-engagement zones.
• Correlated moral valence with protest activity data to assess predictive power of moral framing on mobilization.
• Used statistical modeling to examine how demographic and geographic factors influence moral valence in online discourse.

Experimental results

Research questions

• RQ1 How is moral valence distributed across geographic regions in Los Angeles following the George Floyd incident?
• RQ2 To what extent does moral framing in tweets predict the occurrence and intensity of local protests?
• RQ3 Which moral foundations (e.g., fairness, harm, authority) are most salient in protest-related Twitter discourse in this context?
• RQ4 How do demographic and spatial factors influence the moral tone of online activism in the BLM movement?
• RQ5 What is the relationship between moral valence in social media content and real-world protest mobilization?

Key findings

• Moral valence in tweets was significantly higher in neighborhoods with higher protest activity, indicating a strong link between moral framing and mobilization.
• The moral foundation of 'fairness' was the most frequently cited, followed by 'harm' and 'betrayal', reflecting core concerns about racial injustice and police violence.
• Geographic variation in moral valence was observed, with higher moral valence in predominantly Black and low-income neighborhoods.
• Tweets with higher moral valence were more likely to be retweeted and shared, amplifying their reach and influence.
• Temporal analysis showed spikes in moral valence corresponding to key protest events and media coverage.
• The study confirms that moral discourse on social media is not only reflective but also predictive of protest behavior.