Computational Content Analysis in Advertising Research

Abstract

Computational content analysis (CCA) has experienced a surge in popularity in the field of advertising research. Despite advancements, a comprehensive methodology guide in this area is lacking, presenting challenges for researchers seeking to incorporate these techniques into their study design. This methodology paper aims to provide a thorough overview of CCA applied to different and multiple modalities, including text, images, audio, and video, as a guide for interested researchers. We outline the use of machine learning through CCA in advertising research, covering a wide range of supervised (classification, object detection, emotion analysis, audio sentiment analysis, regression) and unsupervised (topic modeling and clustering) machine learning methods, alongside conventional CCA methods (entity extraction and sentiment analysis). Additionally, we provide a future research agenda that demonstrates how researchers can utilize generative artificial intelligence in CCA.