Online learning has significantly contributed to promoting educational equity, but learners often lack sufficient knowledge assistance and emotional support. While individualized comments (i-Comments) have proven effective in enhancing interaction, manual generation is labor-intensive and difficult to scale. This study proposes a framework for generating i-Comments using ChatGPT based on scaffolding theory and conducts a comprehensive assessment of content quality. Through systematic analysis of sentiment distribution, thematic relevance, and linguistic features, we evaluate the technical feasibility of the content of ChatGPT-generated i-Comments. Results demonstrate that ChatGPT-generated i-Comments achieve 97% positive or neutral emotional tone in emotional support comments and maintain thematic relevance with learning content. Word frequency usage patterns reveal that emotional support employs evaluative language to engage learners affectively; knowledge assistance utilizes technical terminology for precise information transmission, thereby offering insights for future optimization. This study assessed the features of i-Comments content generated by ChatGPT and experimentally validated the effectiveness of ChatGPT-generated i-Comments.