Directed Social Regard: Surfacing Targeted Advocacy, Opposition, Aid, Harms, and Victimization in Online Media
This work provides a more nuanced sentiment analysis tool for social scientists and NLP researchers studying online discourse, though it is an incremental improvement over existing span-level sentiment methods.
The authors propose Directed Social Regard (DSR), a multi-dimensional sentiment analysis approach that detects span-level targets and scores them along three axes of regard, addressing the limitation of traditional sentiment analysis that cannot capture coexisting positive and negative sentiments directed at different targets. They validate their transformer-based models on constructed datasets and report meaningful correlations with labels and topics in six third-party social science datasets.
The language in online platforms, influence operations, and political rhetoric frequently directs a mix of pro-social sentiment (e.g., advocacy, helpfulness, compassion) and anti-social sentiment (e.g., threats, opposition, blame) at different topics, all in the same message. While many natural language processing (NLP) tools classify or score a text's overall sentiment as positive, neutral, or negative, these tools cannot report that positive and negative sentiments coexist, and they cannot report the target of those sentiments. This paper presents the Directed Social Regard (DSR) approach to multi-dimensional, multi-valence sentiment analysis, comprised of a pair of transformer-based models that (1) detects span-level targets of sentiment in a message and then (2) scores all spans within the message context along three (-1, 1) axes of regard that are motivated by social science theories of moral disengagement and moral framing. We present a data collection and annotation strategy for DSR dataset construction, a transformer-based architecture for span-level scoring, and a validation study with promising results. We apply the validated DSR model on six third-party datasets of online media and report meaningful correlations between DSR outputs and the labels and topics in these pre-existing social science datasets.