Understanding Twitter conversations about artificial intelligence in advertising based on natural language processing

Abstract Artificial intelligence (AI) has been widely applied in the advertising industry and attracted increasing attention from advertising scholars. However, the general public’s perceptions of AI in advertising have been relatively neglected within the current literature. In this study, we analyzed people’s posts on Twitter about AI in advertising using natural language processing. Specifically, we conducted topic modeling and sentiment analysis in the Python environment to identify the most salient topics on this issue and the sentiment of these topics. We discovered that what people posted on Twitter about AI in advertising fell into eight different clusters, with the most positive topic being “AI-powered marketing tools” and the most negative topic being “AI’s involvement in social media campaigns.” We believe the findings of this study provide meaningful implications for academic research of AI advertising as well as the implementation of AI in advertising practice.