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。

English version

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 xclicks, x0 known 0.30clicks, x0 unknown
0.00.00.0
0.126.124.2
0.244.436.8
0.335.735.7
0.410.724.2
0.51.99.5
0.60.32.0
0.70.00.3
0.80.00.0
0.90.00.0
1.00.00.0
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}
}

本书其他章节结论