关于「义近与三维打分:加权平均会稀释单维撞车」,《品牌撞名三维预检》给出的定量结论是什么?

结论:第五章加入义近,说明加权平均会稀释单维撞车,示意七对中有 5 对只有取最大值才能标出。

数据口径:在一个示意模型中,声音通道占八成时音码的 AUC 为 0.851,高于字母编辑距离的 0.817。 书中参数为示意;全部计算由书中程序生成;本书是非个人化法律建议,商标近似判断以各地官方检索与专业代理意见为准。

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。

name aname bmeaning distance
yue4_H+yue4_Mjoy+moon0.200
xing1_S+yang2_Wstar+sea0.200
qing1_W+yue4_Hqing2_S+le40.200
qing1_W+yue4_Hming2_S+huan10.600
clara+joyclear+glee0.200
qing1_W+yue4_Hqing4_G+yue4_M1.000
groups 7, tokens 23, near cost 0.2
查看完整验算程序
# Chapter 5. p51: semantic distance from a small synonym table.
# Model: a name is a sequence of meaning tokens (character ids from
# p41, or English words). Tokens in the same synonym group are 0.2
# apart, identical tokens 0, all others 1; the name distance is the
# edit distance over tokens divided by the longer length. The group
# table is an illustrative sample; it includes translation pairs, so
# a Chinese name and an English name can be compared on meaning.
GROUPS = {
    "joy": ("yue4_H", "le4", "huan1", "joy", "glee"),
    "moon": ("yue4_M", "moon", "luna"),
    "star": ("xing1_S", "star", "stella"),
    "clear": ("qing1_W", "qing2_S", "clear", "clara"),
    "bright": ("ming2_S", "bright", "shine"),
    "ocean": ("yang2_W", "ocean", "sea"),
    "festive": ("qing4_G", "fiesta"),
}
TAG = {tok: g for g in sorted(GROUPS) for tok in GROUPS[g]}
NEAR = 0.2


def tdist(a, b):
    if a == b:
        return 0.0
    ga, gb = TAG.get(a), TAG.get(b)
    return NEAR if ga is not None and ga == gb else 1.0


def mdist(a, b):
    prev = [float(j) for j in range(len(b) + 1)]
    for i, x in enumerate(a, 1):
        cur = [float(i)]
        for j, y in enumerate(b, 1):
            cur.append(min(prev[j] + 1.0, cur[j - 1] + 1.0,
                           prev[j - 1] + tdist(x, y)))
        prev = cur
    return prev[-1] / max(len(a), len(b))


PAIRS = (
    (("yue4_H", "yue4_M"), ("joy", "moon")),
    (("xing1_S", "yang2_W"), ("star", "sea")),
    (("qing1_W", "yue4_H"), ("qing2_S", "le4")),
    (("qing1_W", "yue4_H"), ("ming2_S", "huan1")),
    (("clara", "joy"), ("clear", "glee")),
    (("qing1_W", "yue4_H"), ("qing4_G", "yue4_M")),
)

if __name__ == "__main__":
    print("name a              name b             meaning distance")
    for a, b in PAIRS:
        print("{:19s} {:19s} {:8.3f}".format("+".join(a), "+".join(b),
                                             mdist(a, b)))
    print("groups {}, tokens {}, near cost {}".format(
        len(GROUPS), len(TAG), NEAR))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《商标音近形近义近三维距离、从反切到音码的品牌撞名预检母典》第 5 章「义近与三维打分:加权平均会稀释单维撞车」(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-317 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026317,
  author    = {Fan, Yafu},
  title     = {Phonetic, Graphic and Semantic Distance between Trademarks: Pre-Screening Brand-Name Collisions from Fanqie Spellings to Sound Codes},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论