Artificial Intelligence and Inclusion: Formerly Gang-Involved Youth as Domain Experts for Analyzing Unstructured Twitter Data

Abstract

Mining social media data for studying the human condition has created new and unique challenges. When analyzing social media data from marginalized communities, algorithms lack the ability to accurately interpret off-line context, which may lead to dangerous assumptions about and implications for marginalized communities. To combat this challenge, we hired formerly gang-involved young people as domain experts for contextualizing social media data in order to create inclusive, community-informed algorithms. Utilizing data from the Gang Intervention and Computer Science Project—a comprehensive analysis of Twitter data from gang-involved youth in Chicago—we describe the process of involving formerly gang-involved young people in developing a new part-of-speech tagger and content classifier for a prototype natural language processing system that detects aggression and loss in Twitter data. We argue that involving young people as domain experts leads to more robust understandings of context, including localized language, culture, and events. These insights could change how data scientists approach the development of corpora and algorithms that affect people in marginalized communities and who to involve in that process. We offer a contextually driven interdisciplinary approach between social work and data science that integrates domain insights into the training of qualitative annotators and the production of algorithms for positive social impact.

When Twitter Fingers Turn to Trigger Fingers: a Qualitative Study of Social Media-Related Gang Violence

“…This study could begin important conversations between violence prevention organizations, schools, and law enforcement for considering social media as a tool for the prevention and intervention of violence. To be sure, interpreting tweets is not an exact science but enlisting the active assistance of young people in the process (Frey et al 2018) is an important step in contextualizing content and identifying the practice of how the offline environment and experience shape social media communication in ways that may lead to violence.…”

“…In this study, we offer an alternative methodological approach to studying the implications of social media for gangs: interpretations of social media posts by Bdomain experts ( Frey et al 2018). We confront the idea that language used by Black and Latino youth who are gang-involved on Twitter can consist of a mix of African American Vernacular English, short acronyms frequently used on social media and hyperlocal contextual mentions of institutions, events, and experiences (Chang et al n.d.).…”

“…Recent research suggests that the racial decoding of social media text of Black individuals accused of a crime may be misinterpreted and monitored by law enforcement whereas the same strategies and interpretations may not occur with perpetrators of crimes who happen to be white (Patton et al 2017d). The aforementioned strategies, while strong independently, were not sufficient for a hyper local and contextually driven analysis of social media communication among gang and non-gang involved boys and men who are Black and Latino because they lacked domain expertise, specifically involvement of gang-involved youth and local community members in the interpretation and translation of social media text (Frey et al 2018). As such, we designed a mixed method methodological strategy (e.g., qualitative interviews and quantitative coding scheme) that privileges the interpretations and domain expertise of Chicago Black and Latino youth who, at the time of the study, claimed current or former gang experiences.…”

“…We used a quantitative coding scheme to compare perceptions of tweets across the participants. Specifically, the article leverages the expertise of formally gang-involved youth as domain experts to interpret and categorize potential threats online (Frey et al 2018), using their responses to develop a framework for understanding participant responses to tweets and the potential for violence in tweets. As such, this article makes two important contributions.…”

Youth and Digital Citizenship+ (Plus): Understanding Skills for a Digital World

“…In light of the importance of context, it may also be useful for young people to recognize that while AI systems can have impacts on a global scale, the relatively small group of companies leading the development of AI-powered technologies typically have little insight from a variety of disciplines, cultures, and socioeconomic classes (Ashar & Cortesi, 2018), underscoring the importance of including underrepresented groups in the design and deployment of AI (Frey, Patton, Gaskell, & McGregor, 2020). Against this backdrop, there is a growing gap between those who have access to information about AI systems and understand their impact and those who do not (Ashar & Cortesi, 2018).…”