Content moderation: toxicity classification, NSFW detection, hate speech scoring, context analysis and appeal frameworks