关于「内链拓扑:PageRank 漏斗与询盘页的稳态概率」,《长尾页信息增益与收录排程》给出的定量结论是什么?

结论:第四章比较四种内链拓扑,漏斗把询盘页推到均匀分布的 7.12 倍,访客平均 2.84 次点击到达。

数据口径:书中参数为示意,全部计算由书中程序生成。 本书是非个人化经营建议;房产内容不构成法律、税务或投资建议,塔罗部分不做个人占断,宗祠资料须经本人同意,书中数据全部合成。

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 标准库与结论吻合的打印值: 7.12

topologycheckoutx uniformhubstailshome
mesh0.02861.000.08570.85710.0286
islands0.00630.220.01900.96500.0097
hub-spoke0.01540.540.45330.47880.0524
funnel0.20337.120.27600.32400.1967
funnel total mass check 1.000000
查看完整验算程序
# Chapter 4. p41: where internal PageRank settles under four link
# topologies (Brin and Page 1998; Page et al. 1999). A 35-page site:
# home (0), three topic hubs (1-3), ten long-tail pages per hub (4-33)
# and one inquiry page, the "checkout" (34). Power iteration with
# damping 0.85; dangling pages spread their mass evenly. This is the
# textbook random-surfer model of internal links only, not a ranking
# formula used by any search engine today.
import math

N, HOME, CHECK = 35, 0, 34
HUBS = [1, 2, 3]


def tails(h):
    return list(range(4 + 10 * (h - 1), 14 + 10 * (h - 1)))


def build(kind):
    out = {i: [] for i in range(N)}
    if kind == "mesh":
        for i in range(N):
            out[i] = [j for j in range(N) if j != i]
        return out
    out[HOME] = HUBS + [CHECK]
    out[CHECK] = [HOME]
    for h in HUBS:
        ts = tails(h)
        if kind == "islands":
            out[h] = ts
            for t in ts:
                out[t] = [x for x in ts if x != t]
        elif kind == "hub-spoke":
            out[h] = ts + [HOME]
            for t in ts:
                out[t] = [h]
        elif kind == "funnel":
            out[h] = ts + [CHECK, HOME]
            for t in ts:
                out[t] = [h, CHECK]
    return out


def pagerank(out, d=0.85, iters=200):
    pr = [1.0 / N] * N
    for _ in range(iters):
        dang = math.fsum(pr[i] for i in range(N) if not out[i])
        inflow = [[] for _ in range(N)]
        for i in range(N):
            for j in out[i]:
                inflow[j].append(pr[i] / len(out[i]))
        pr = [(1 - d) / N + d * (math.fsum(inflow[j]) + dang / N)
              for j in range(N)]
    return pr


if __name__ == "__main__":
    print("topology    checkout  x uniform   hubs    tails    home")
    for kind in ("mesh", "islands", "hub-spoke", "funnel"):
        pr = pagerank(build(kind))
        hubs = math.fsum(pr[h] for h in HUBS)
        tl = math.fsum(pr[4:34])
        print("{:<10}  {:.4f}    {:>5.2f}    {:.4f}  {:.4f}  {:.4f}".format(
            kind, pr[CHECK], pr[CHECK] * N, hubs, tl, pr[HOME]))
    pr = pagerank(build("funnel"))
    print("funnel total mass check {:.6f}".format(math.fsum(pr)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《多域名程序化SEO长尾拓扑、新站沙盒极速收录与高意图自然流量收割母典》第 4 章「内链拓扑:PageRank 漏斗与询盘页的稳态概率」(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-346 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026346,
  author    = {Fan, Yafu},
  title     = {Programmatic SEO Long-Tail Topology, Rapid Indexation and High-Intent Organic Traffic Harvesting across Multi-Domain Fleets},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论