Our paper titled “Automatically Detecting Image–Text Mismatch on Instagram with Deep Learning” is published at Journal of Advertising, which is ranked at the top rank in Communication (IF=6.302). To address the usage of brand-relevant hashtags for promoting brand-irrelevant content, we propose to use deep learning for the detection of such brand-incongruent content from three cues: image, text, and metadata. We also discuss the advantages of machine-learning methods as a novel research tool for advertising research.
This study was done with the interdisciplinary team of communication/advertising scholars and computers scientists. You can check the open-access article at [link]!