Chinese Personal Name Variants

The number of Chinese personal names and their variants is very large (in the millions) and identifying those is a difficult computational linguistic task. To enhance named-entity recognition (NER) technology and various other NLP applications, CJKI maintains a comprehensive dataset of over 1.6 million Chinese seed names (surnames and given names separately) and approximately 7.6 million romanized variants for these names.

CJKI’s Database of Chinese Personal Name Variants (CNV) includes Chinese first names and last names in all the standard romanization systems. CNV also includes classification codes, frequency of occurrence statistics, and gender codes. There are many systems for romanizing Chinese, such as Hanyu Pinyin, Wade-Giles, Yale, and Tongyong Pinyin (Taiwan), as well as various other popular romanization systems, and many ways to romanize Chinese that have fallen out of use.

CNV provides comprehensive coverage for the major Chinese romanization systems and their variants. We also maintain a Chinese Name Dialectal Variants database for dialectical romanized variants, covering Cantonese, Hakka and Hokkien dialects.