An Annotated Image Dataset from Facebook Post Comments for Social Appropriateness and Humor Analysis

This dataset contains images collected from publicly available Facebook post comments. Each image was manually annotated into one of three categories—Good, Bad and Funny based on contextual and social considerations. Images are organized into class-specific folders, with sequential numeric file names used for indexing. The dataset is intended for research in social media analysis, content moderation, humor detection, and computer vision.Steps to reproduce1. Download and extract the dataset from Mendeley Data. 2. The dataset contains three folders: Good, Bad, and Funny. 3. Each folder includes images belonging to that category. 4. Use the folder names as labels for analysis or model training.