Abstract
<p>在社会语言学与文字政策交叉领域，《通用规范汉字表》附表及GB/T 12345-90虽对简繁“一对多”字际关系有所整理，但均存在收字不全、术语混编的问题。为弥补上述缺漏，本文编制了《简繁一对多155组繁本字对照表》。该表以官方96组为种子集，经比对多部字书及区域标准，补充27组区域差异字，筛得同形异音字21组，并经语料共现校验，最终闭合为155组。编制中引入“繁本字”概念（繁本字 = 繁体字 ∪ 传承字），以“四重意涵”（文本之本、本字之本、本有之字之本、文化本位之本）与“三阶证成”（语形、语义、语用）为理论根基，厘清了简化字与多个繁本字之间的音义对应关系。基于0.97亿字语料的简繁转换实验证明，以“繁本字+155组音义规则”为目标层，可将字符级错误率由17.5ppm降至2.8ppm，降幅达84%。本表可供简繁转换软件开发、辞书编纂、汉字教学及书写实践参考使用，也可为未来字表修订提供参考。英文摘要Abstract:The appendices of the Table of General Standard Chinese Characters and GB/T 12345-90 have addressed the "one-to-many" relationships between simplified and traditional characters, but both suffer from incomplete coverage and terminological inconsistencies. To address these gaps, this paper presents the One-to-Many 155 Groups Fanbenzi Mapping Table. Starting from the official 96 groups as the seed set, the table was expanded by cross-referencing multiple dictionaries and regional standards, adding 27 groups of regionally variant characters and 21 groups of homographs with different pronunciations, and finally closed at 155 groups through corpus-based co-occurrence validation.The compilation is grounded in the concept of "Fanbenzi" (Fanbenzi = Traditional Characters ∪ Inherited Characters), supported by its "Four Meanings" and "Three Justifications" as theoretical foundations. Each simplified character's corresponding Fanbenzi variants are distinguished by phonetic and semantic contexts. In a simplified-to-traditional conversion experiment conducted on a 97-million-character corpus, using "Fanbenzi + 155 phonetic-semantic rules" as the target layer reduced the character-level error rate from 17.5 ppm to 2.8 ppm, a reduction of 84%. This table may serve as a reference for simplified-traditional conversion software development, lexicography, Chinese character teaching, and calligraphy practice, as well as for future revisions of official character tables.</p>