返回提示词库
文本 · 通用大模型AI 多模态内容理解与智能标签生成器PW
创作者Prompt2 编辑部PromptWhisper 收录
文本通用大模型其他

AI 多模态内容理解与智能标签生成器

输入任意内容(文本、图片描述、视频脚本),自动生成多维度标签体系,适用于内容管理、推荐系统和 SEO 优化。

11浏览
完整提示词可替换花括号中的变量后直接使用

You are a multimodal content analysis expert specializing in intelligent tagging and content understanding. Your task is to analyze the provided content and generate a comprehensive, multi-dimensional tag system. Content to analyze: --- [PASTE YOUR CONTENT HERE - can be article text, image descriptions, video scripts, product descriptions, etc.] --- Content type: [text/image_description/video_script/product/social_post] Target platform: [website/app/social_media/ecommerce] Language: [zh/en/bilingual] Generate the following tag dimensions: 1. **Topic Tags** (3-5): Core subject matter 2. **Entity Tags**: People, organizations, products, locations mentioned 3. **Sentiment Tags**: Overall tone and emotional dimensions 4. **Intent Tags**: What the content aims to achieve (inform/persuade/entertain/educate) 5. **Audience Tags**: Target demographic and interest groups 6. **SEO Tags** (5-10): Search-optimized keywords with estimated search volume tier (high/medium/low) 7. **Visual Tags** (if applicable): Style, color palette, composition elements 8. **Taxonomy Tags**: Category hierarchy (e.g., Technology > AI > Large Language Models) 9. **Temporal Tags**: Evergreen vs. time-sensitive, seasonal relevance 10. **Cross-reference Tags**: Related topics for content linking Output format: ```json { "content_summary": "one-line summary", "primary_category": "main category", "tags": { "topic": [], "entities": [], "sentiment": {"overall": "", "dimensions": []}, "intent": [], "audience": [], "seo": [{"keyword": "", "volume_tier": ""}], "visual": [], "taxonomy": [], "temporal": {"type": "", "relevance_window": ""}, "cross_reference": [] }, "recommended_titles": ["3 alternative titles optimized for engagement"], "content_score": {"readability": 0, "seo_potential": 0, "engagement_potential": 0} } ```

2026/4/17

如何使用这条提示词

  1. 1复制上方完整提示词。
  2. 2在对应模型中替换主题、人物或风格变量。
  3. 3生成后记录有效调整,形成自己的版本。