A Framework for Generating Images and Hashtags for Social Media Posts for Artificial Influencers

Abstract

This paper introduces a novel framework that leverages artificial intelligence to create engaging social media content, encompassing both images and hashtags, using a diffusion model and interactive evolution computation. In the era of digital influencers, this approach stands out by not only generating new content that adapts to user feedback but also by its capability to engage audiences effectively. Through a series of experiments, the framework’s performance was assessed based on the quality, diversity, and engagement level of the content it produces. The evaluation was conducted using both offline metrics and real-world online testing on Instagram, revealing that the framework not only excels in fostering user engagement in controlled environments but also sustains this engagement in live social media settings. The findings underscore the promising application of artificial influencers in content generation and offer valuable insights into the deployment of the proposed framework in this burgeoning field.