关于「Facebook:小组、本地信任与弱连接,以及经由朋友网络的触达」,《五大平台算法与一鱼五吃》给出的定量结论是什么?

结论:第五章说明种子放在跨圈者身上触达 15.36 人,同圈种子只有 7.89 人。

数据口径:书中把每条平台说法标为三类:带日期的官方披露、同行评审研究、本书的示意模型;书中的权重、概率与漏斗参数均为示意,不是任何平台的真实设定,全部计算由书中程序生成。 本书是一般性经营参考,不构成法律、税务或投资建议;不涉及假账号、机器人、互赞群、买互动、抓取个人资料或群发私讯;合作内容一律标明;达人酬劳只有单层直付,不涉及任何多层分润。

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: 15.36, 7.89

240 people, 705 ties inside clusters, 100 bridging ties
one clustermean reach7.89 people, clusters touched 1.60
three clustersmean reach10.63 people, clusters touched 3.51
bridge peoplemean reach15.36 people, clusters touched 4.14
查看完整验算程序
# Chapter 5. p51: where to seed a local post in a friend network of six clusters.
# Six clusters of 40 people (estates, schools, clubs); a tie inside a cluster exists
# with probability 0.15, between clusters with 0.004 (seeded random graph). A share
# passes along a tie with probability 0.12 (independent cascade). Three seeds are
# placed (a) in one cluster, (b) one each in three clusters, (c) on the three people
# with most cross-cluster ties. 500 seeded runs each. All parameters are illustrative.
import math
import random

RNG = random.Random(1973)
K, M, P_IN, P_OUT, P_SHARE, RUNS = 6, 40, 0.15, 0.004, 0.12, 500
N = K * M
adj = [[] for _ in range(N)]
for i in range(N):
    for j in range(i + 1, N):
        same = i // M == j // M
        if RNG.random() < (P_IN if same else P_OUT):
            adj[i].append(j)
            adj[j].append(i)
cross = [len([j for j in adj[i] if j // M != i // M]) for i in range(N)]
ties_in = len([1 for i in range(N) for j in adj[i] if j > i and j // M == i // M])
ties_out = len([1 for i in range(N) for j in adj[i] if j > i and j // M != i // M])
print("%d people, %d ties inside clusters, %d bridging ties" % (N, ties_in, ties_out))


def spread(seeds):
    active, frontier = set(seeds), list(seeds)
    while frontier:
        nxt = []
        for u in frontier:
            for v in adj[u]:
                if v not in active and RNG.random() < P_SHARE:
                    active.add(v)
                    nxt.append(v)
        frontier = nxt
    return len(active), len({u // M for u in active})


plans = {"one cluster": [0, 1, 2], "three clusters": [0, M, 2 * M],
         "bridge people": sorted(range(N), key=lambda i: (-cross[i], i))[:3]}
for name, seeds in plans.items():
    reach, clus = [], []
    for _ in range(RUNS):
        r, c = spread(seeds)
        reach.append(r)
        clus.append(c)
    print("%-15s mean reach %6.2f people, clusters touched %.2f" % (
        name, math.fsum(reach) / RUNS, math.fsum(clus) / RUNS))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《五大社交图谱算法异质性、跨平台同源异构分发与达人聚拢母典》第 5 章「Facebook:小组、本地信任与弱连接,以及经由朋友网络的触达」(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-359 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026359,
  author    = {Fan, Yafu},
  title     = {Algorithmic Heterogeneity across RedNote, Threads, X, Instagram and Facebook: Cross-Platform Adaptation and Creator Aggregation},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论