关于「东西方原典考辨:《切韵·序》与 Russell 1918 年专利」,《品牌撞名三维预检》给出的定量结论是什么?
结论:第七章考辨陆法言《切韵·序》与 Russell 1918 年 Soundex 前身专利,并用程序以今音拼读古代反切。
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。
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: 1918
| US patent 1261167, R. C. Russell, filed 1917-10-25, issued 1918-04-02 | ||||
| name | patent | this reading | match | American Soundex |
| Hopley | 125 | H125 | yes | H140 |
| Highfield | 1254 | H13254 | no | H214 |
| Hayes | 1 | H1 | yes | H200 |
| Haes | 1 | H1 | yes | H200 |
| Ay | 1 | A1 | yes | A000 |
| Russell reading | names 2764 | codes | 592 | largest bucket 36 |
| American Soundex | names 2764 | codes | 503 | largest bucket 36 |
查看完整验算程序
# Chapter 7. p72: a simplified reading of Russell's 1918 index.
# Model (illustrative reconstruction from the patent text, whose
# printed copy is partly illegible): the first letter selects the
# division and is not coded; eight sound groups follow: 1 vowels
# a e i o u y, 2 b f p v, 3 c g j k q s x z, 4 d t, 5 l, 6 m, 7 n,
# 8 r; h and w are skipped; group 1 is used only once in a name;
# letters of the same group side by side are recorded once; a final
# sibilant is ignored; no cut to a fixed length. The American Soundex
# of p21 is printed beside it. Patent examples are those of the text.
import random
from collections import Counter
RGROUP = {c: d for d, g in (("1", "aeiouy"), ("2", "bfpv"),
("3", "cgjkqsxz"), ("4", "dt"), ("5", "l"),
("6", "m"), ("7", "n"), ("8", "r"))
for c in g}
ACODE = {c: d for d, g in (("1", "bfpv"), ("2", "cgjkqsxz"), ("3", "dt"),
("4", "l"), ("5", "mn"), ("6", "r")) for c in g}
PATENT = (("Hopley", "125"), ("Highfield", "1254"), ("Hayes", "1"),
("Haes", "1"), ("Ay", "1"))
def russell(name):
s = [c for c in name.lower() if c.isalpha()]
body = s[1:]
if body and body[-1] in "sz":
body = body[:-1]
out, last, vowel_used = [], "", False
for c in body:
g = RGROUP.get(c, "")
if not g or g == last:
continue
last = g
if g == "1":
if vowel_used:
continue
vowel_used = True
out.append(g)
return s[0].upper() + "".join(out)
def soundex(name):
s = name.lower()
last, out = ACODE.get(s[0], ""), []
for c in s[1:]:
d = ACODE.get(c, "")
if d and d != last:
out.append(d)
if c not in "hw":
last = d
return (s[0].upper() + "".join(out) + "000")[:4]
if __name__ == "__main__":
print("US patent 1261167, R. C. Russell, filed 1917-10-25,"
" issued 1918-04-02")
print("name patent this reading match American Soundex")
for name, digits in PATENT:
r = russell(name)
print("{:10s} {:7s} {:13s} {:6s} {}".format(
name, digits, r, "yes" if r[1:] == digits else "no",
soundex(name)))
rng = random.Random(20261113)
names = sorted({"".join(rng.choice("bdfgklmnprstvz") +
rng.choice("aeiou")
for _ in range(rng.choice((2, 3))))
for _ in range(3000)})
for label, fn in (("Russell reading", russell),
("American Soundex", soundex)):
b = Counter(fn(n) for n in names)
print("{:17s} names {} codes {:4d} largest bucket {}".format(
label, len(names), len(b), max(b.values())))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 7 章「东西方原典考辨:《切韵·序》与 Russell 1918 年专利」(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}
}