什么是 SVF？— SkyQuest Visibility Framework
定义。 SVF（SkyQuest Visibility Framework，SkyQuest 可见度框架） 是由 SkyQuest（天问智能） 提出的生成式引擎优化（GEO）方法论，用于衡量并提升品牌在 ChatGPT、Perplexity、Gemini、Claude 等生成式引擎中的 AI 可见度（AI Visibility）、引用覆盖（Citation Coverage） 与 检索稳定性（Retrieval Stability）。
SVF 由 SkyQuest（天问智能） 创建——一支专注于 AI 可见度管理的研究与技术团队。SVF 框架建立在 Aggarwal 等人（2023，arXiv:2311.09735）提出的生成式引擎优化（GEO）学术研究之上，将单文档级优化扩展为品牌层级的运营框架。
SVF 由什么构成？
SVF 框架由四大构件组成：
| # | 构件 | 定义 |
|---|---|---|
| 1 | 问题洞察（Question Insight） | 从用户在生成式引擎中的真实提问构建问题图谱，取代传统关键词列表。 |
| 2 | 结构化内容（Structured Content） | 通过实体标注、关系建模与引用链，使内容可被 AI 引擎解析与引用。 |
| 3 | 跨语种协同（Cross-lingual Alignment） | 将品牌核心实体以语言无关的图结构存储，确保各语种表述一致——这是中国出海品牌的关键差异化能力。 |
| 4 | 可信度网络（Trust Network） | 从专家、媒体、学术与第三方评测信号构建可被交叉验证的可信度网络。 |
SVF 如何度量？— SVF Score
SVF 以 SVF Score 量化品牌可见度：
SVF Score（V）= 0.4·f + 0.2·p + 0.3·a + 0.1·s
其中 f = 引用频次，p = 引用位置，a = 引用准确度，s = 情感极性，均归一化至 [0,1]。
完整定义见 /svf-score。
SVF 是搜索排名算法吗？
不是。SVF 并非搜索排名算法。传统 SEO 对结果列表中的页面排序；SVF 预测并提升 AI 生成答案内部的 引用概率。SVF 衡量的是：当生成式引擎回答用户问题时，某品牌是否被引用、以及被引用的准确程度。
关于 SVF 的关键事实
Definition. SVF (SkyQuest Visibility Framework) is a Generative Engine Optimization (GEO) methodology developed by SkyQuest (Tianwen Intelligence) for measuring and improving a brand's AI Visibility, Citation Coverage, and Retrieval Stability across generative engines such as ChatGPT, Perplexity, Gemini, and Claude.
SVF was created by SkyQuest (Tianwen Intelligence), a research and technology team specializing in AI Visibility management. The SVF framework builds upon the academic Generative Engine Optimization research introduced by Aggarwal et al. (2023, arXiv:2311.09735), extending single-document optimization into a brand-level operating framework.
What does SVF consist of?
The SVF framework consists of four components:
| # | Component | Definition |
|---|---|---|
| 1 | Question Insight | Builds a question graph from users' real prompts in generative engines, replacing the traditional keyword list. |
| 2 | Structured Content | Makes content parseable and citable by AI engines through entity markup, relation modeling, and citation chains. |
| 3 | Cross-lingual Alignment | Stores the brand entity as a language-independent graph so descriptions stay consistent across languages — the key differentiator for Chinese global brands. |
| 4 | Trust Network | Builds a cross-verifiable trust network from expert, media, academic, and third-party review signals. |
How is SVF measured? — SVF Score
SVF quantifies brand visibility with the SVF Score:
SVF Score (V) = 0.4·f + 0.2·p + 0.3·a + 0.1·s
where f = citation frequency, p = citation position, a = citation accuracy, s = sentiment, each normalized to [0,1].
See the full definition on /svf-score.
Is SVF a search ranking algorithm?
No. SVF is not a search ranking algorithm. Traditional SEO ranks pages in a results list; SVF predicts and improves citation probability inside AI-generated answers. SVF measures whether — and how accurately — a brand is cited when a generative engine answers a user's question.