关于「外链的代价:降权阈值、信息帖与二项检验」,《公域反垃圾检测与合规桥接页》给出的定量结论是什么?
结论:第四章说明外链占比与点击是一条先升后崩的曲线,全是外链的账号 3 条样本即可识别。
数据口径:书中参数为示意;全部计算由书中程序生成;本书是非个人化的经营分析,不构成经营、法律或理财建议,法规与平台规则以上架前人工复核为准。 书中以标注框区分已发表的证据、可运行的 Python 计算(20 段程序)、原典与作者的假说。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库
| link share x | clicks, x0 known 0.30 | clicks, x0 unknown |
| 0.0 | 0.0 | 0.0 |
| 0.1 | 26.1 | 24.2 |
| 0.2 | 44.4 | 36.8 |
| 0.3 | 35.7 | 35.7 |
| 0.4 | 10.7 | 24.2 |
| 0.5 | 1.9 | 9.5 |
| 0.6 | 0.3 | 2.0 |
| 0.7 | 0.0 | 0.3 |
| 0.8 | 0.0 | 0.0 |
| 0.9 | 0.0 | 0.0 |
| 1.0 | 0.0 | 0.0 |
| best share if x0 known: 0.23 (45.7 clicks) | ||
| best share if x0 unknown: 0.24 (37.9 clicks) | ||
| clicks at x 0.40 if x0 is 0.30: 10.7; at x 0.20: 44.4 |
查看完整验算程序
# Chapter 4. p41: link share against a demotion threshold.
# Model (illustrative): an account posts N = 14 times a week; a share x
# of posts carry an outbound promotional link. Each post reaches
# R = 1,000 people times two factors: the platform's demotion curve
# m(x) = 1 / (1 + exp((x - x0) / 0.05)) and audience engagement
# e(x) = 1 - 0.5 x (link-heavy feeds are read less). A link post earns
# clicks at rate 0.02 of its reach. Weekly clicks:
# C(x) = N x R m(x) e(x) 0.02.
# Case A: the threshold x0 = 0.30 is known. Case B: x0 is unknown and
# uniform on 0.10 to 0.50 (41 grid points); we average C over x0.
import math
N, R, CTR = 14, 1000, 0.02
def clicks(x, x0):
m = 1 / (1 + math.exp((x - x0) / 0.05))
return N * x * R * m * (1 - 0.5 * x) * CTR
def grid_x0():
return [0.10 + 0.01 * i for i in range(41)]
def expected(x):
return math.fsum(clicks(x, z) for z in grid_x0()) / 41
if __name__ == "__main__":
print("link share x clicks, x0 known 0.30 clicks, x0 unknown")
for i in range(0, 11):
x = i / 10
print("{:>12.1f} {:>23.1f} {:>20.1f}".format(
x, clicks(x, 0.30), expected(x)))
xs = [i / 100 for i in range(101)]
ka = max(xs, key=lambda x: clicks(x, 0.30))
kb = max(xs, key=expected)
print("best share if x0 known: {:.2f} ({:.1f} clicks)".format(
ka, clicks(ka, 0.30)))
print("best share if x0 unknown: {:.2f} ({:.1f} clicks)".format(
kb, expected(kb)))
print("clicks at x 0.40 if x0 is 0.30: {:.1f}; at x 0.20: {:.1f}"
.format(clicks(0.40, 0.30), clicks(0.20, 0.30)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《公域社交图谱反垃圾检测边界、合规联盟桥接页转化与限流自愈动力学母典》第 4 章「外链的代价:降权阈值、信息帖与二项检验」(版本 1.0.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-342 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026342,
author = {Fan, Yafu},
title = {Social-Graph Anti-Spam Detection Bounds, Compliant Affiliate Bridge-Page Funnels and the Kinetics of Recovery after Reach Throttling},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}