GEO 与 SEO：从排名到引用的工程化实践

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Article",
      "@id": "https://example.com/blog/geo-seo-practice",
      "headline": "GEO 与 SEO：从排名到引用的工程化实践",
      "description": "探讨生成式引擎优化与传统搜索引擎优化的技术差异与协同策略。",
      "author": {
        "@type": "Person",
        "name": "作者姓名",
        "url": "https://example.com/about",
        "sameAs": [
          "https://github.com/username",
          "https://twitter.com/username"
        ]
      },
      "publisher": {
        "@type": "Organization",
        "name": "站点名称",
        "logo": {
          "@type": "ImageObject",
          "url": "https://example.com/assets/logo.png"
        }
      },
      "datePublished": "2025-08-15T08:00:00+08:00",
      "dateModified": "2025-09-01T14:30:00+08:00",
      "mainEntityOfPage": {
        "@type": "WebPage",
        "@id": "https://example.com/blog/geo-seo-practice"
      }
    },
    {
      "@type": "FAQPage",
      "mainEntity": [
        {
          "@type": "Question",
          "name": "GEO 会取代 SEO 吗？",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "不会。GEO 建立在 SEO 的技术基础之上，两者是互补关系。SEO 确保内容被发现和索引，GEO 确保内容被 AI 系统引用和总结。"
          }
        },
        {
          "@type": "Question",
          "name": "如何衡量 GEO 的效果？",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "核心指标包括品牌在 AI 生成答案中的引用频率、引用上下文的相关性，以及因 AI 引用带来的品牌搜索量增长和辅助转化。"
          }
        }
      ]
    }
  ]
}
</script>

# macOS / Linux
pip install PyLD requests

# Windows (PowerShell)
pip install PyLD requests

import json
from pyld import jsonld

doc = {
    "@context": "https://schema.org",
    "@type": "Article",
    "headline": "测试文章",
    "author": {"@type": "Person", "name": "Test Author"}
}

# 验证基本结构
compacted = jsonld.compact(doc, doc)
print(json.dumps(compacted, indent=2, ensure_ascii=False))

User-agent: GPTBot
Allow: /

User-agent: OAI-SearchBot
Allow: /

User-agent: DeepSeekBot
Allow: /

User-agent: KimiBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: CCBot
Allow: /

# 站点名称

> 一句话描述站点定位。

## 核心内容

- [GEO 与 SEO 实践指南](https://example.com/blog/geo-seo): 生成式优化与传统搜索优化的协同策略
- [Nginx 性能调优手册](https://example.com/blog/nginx-tuning): 从内核参数到应用层的完整配置参考

## 可选内容

- [关于我们](https://example.com/about): 团队背景与联系方式

## 使用政策

允许用于回答用户查询时的实时检索与引用。
不允许用于模型训练或数据挖掘。

import os
import json
import requests
from datetime import datetime

# 配置
API_KEY = os.getenv("OPENAI_API_KEY")
BASE_URL = os.getenv("OPENAI_BASE_URL", "https://api.openai.com/v1")
QUERIES = [
    "如何优化网站 SEO 结构",
    "GEO 与 SEO 有什么区别",
    "最好的 Python Web 框架是什么"
]
BRAND_KEYWORDS = ["example.com", "你的品牌名"]
OUTPUT_FILE = "geo_monitor.json"

def check_citations(query, brand_keywords):
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    payload = {
        "model": "gpt-4o-mini",
        "messages": [
            {"role": "system", "content": "请搜索网络并给出准确、简洁的回答，同时列出你参考的来源链接。"},
            {"role": "user", "content": query}
        ],
        "tools": [{"type": "web_search_preview"}],
        "tool_choice": {"type": "web_search_preview"}
    }

    try:
        resp = requests.post(
            f"{BASE_URL}/chat/completions",
            headers=headers,
            json=payload,
            timeout=60
        )
        data = resp.json()
        content = data["choices"][0]["message"]["content"]

        # 尝试提取引用链接（不同模型格式不同，这里做简化处理）
        citations = []
        message = data["choices"][0]["message"]
        if "annotations" in message:
            citations = [a.get("url", "") for a in message["annotations"] if a.get("type") == "url_citation"]

        # 检查品牌是否在引用链接或正文中出现
        matched = any(
            kw in str(citations) or kw in content
            for kw in brand_keywords
        )
        return {
            "query": query,
            "matched": matched,
            "citation_count": len(citations),
            "sample_text": content[:200]
        }
    except Exception as e:
        return {"query": query, "error": str(e)}

def main():
    results = []
    for q in QUERIES:
        print(f"Checking: {q}")
        results.append(check_citations(q, BRAND_KEYWORDS))

    record = {
        "checked_at": datetime.now().isoformat(),
        "results": results
    }

    # 追加到历史记录
    history = []
    if os.path.exists(OUTPUT_FILE):
        with open(OUTPUT_FILE, "r", encoding="utf-8") as f:
            history = json.load(f)

    history.append(record)
    with open(OUTPUT_FILE, "w", encoding="utf-8") as f:
        json.dump(history, f, ensure_ascii=False, indent=2)

    print(f"Results saved to {OUTPUT_FILE}")

if __name__ == "__main__":
    main()

# 设置环境变量后执行
export OPENAI_API_KEY="your_key"           # macOS/Linux
export OPENAI_BASE_URL="https://your-proxy.com/v1"  # 如需代理

# $env:OPENAI_API_KEY="your_key"           # Windows PowerShell
# $env:OPENAI_BASE_URL="https://your-proxy.com/v1"

python geo_monitor.py

BASE_URL = "https://api.deepseek.com"
# payload 中 model 改为 "deepseek-chat"