Algorithmic Inclusion Patterns / 算法包含模式

Algorithmic inclusion patterns are the choices that determine what content, behavior, user signal, or social reality enters the algorithmic field. In 159.算法的六副面孔：它是如何从处理数据，变成定义我们是谁的, this is the first face of algorithms: before a platform ranks anything, it has already decided what can be indexed, tagged, measured, and used.

The source treats inclusion as political because classification defines what something is and where it is allowed to appear. Its Amazon Fail example is the cleanest case: LGBTQ-related books could lose bestseller visibility after classification as adult content even if they were not removed from Amazon entirely. The same pattern can apply to hot searches, feeds, recommendation pools, and ranking systems where demotion is less visible than deletion.

Key Claims

• Algorithmic power begins at intake: excluded or badly tagged material may become invisible before ranking even starts.
• Tags and fields are not neutral descriptors; they can decide whether a person, group, topic, or work becomes legible to the platform.
• Marginal groups can be underrepresented when they do not generate measurable online signals or when their signals are classified as low-value, risky, or irrelevant.
• Inclusion patterns extend Algorithmic Labeling from social labels into the platform’s own data structure.
• Demotion can govern speech without producing the visible drama of removal.

Connections

• Public Relevance Algorithms / 公共相关性的算法 — umbrella category.
• Algorithmic Labeling — adjacent social compression pattern.
• Algorithmic Relevance Assessment / 算法相关性评估 — later ranking stage affected by what was admitted and labelled.
• Amazon and Google — platform examples where classification and search visibility matter.
• Content Ecosystem Governance, Platform Data Regulation, and AI Information Pollution — governance and trust branches where inclusion/exclusion decisions become high stakes.