A conceptual framework for ideology beyond the left and right
Kenneth Joseph, Kim Williams, David Lazer
Most AI research studying political ideology in text has oversimplified it into a simple left-vs-right spectrum, but people actually hold nuanced views across many different topics like race, climate, and gender. This paper proposes a better framework that treats ideology as a complex network of connected concepts, showing how ideology appears in language through various processes like how issues are framed. By connecting computational AI methods with ideology research from social science, the work opens up new ways to analyze political discourse online.
Natural Language ProcessingComputational Social ScienceIdeology DetectionPolitical Discourse