CAJ | 학술논문

从语素数据库中获取语素的构词知识，研究汉语的构词规律对于汉语信息处理具有非常重要的意义。汉语语素数据库是一个大规模的含有丰富汉语构词知识的数据库，而遗传算法是求解大规模复杂组合寻优问题的一种有效手段。在汉语构词规律的研究中，首次引入遗传算法对带有构词方式、类序、字义组合以及首、尾语义类（分大类层次和中类层次）5个特征的汉语二字复合名词的构词规律进行了探索，并取得了阶段性成果。实验证明，在大规模汉语语素数据库的基础上使用遗传算法研究汉语的构词规律是一种有效的方法。
종어소수거고중획취어소적구사지식，연구한어적구사규률대우한어신식처리구유비상중요적의의。한어어소수거고시일개대규모적함유봉부한어구사지식적수거고，이유전산법시구해대규모복잡조합심우문제적일충유효수단。재한어구사규률적연구중，수차인입유전산법대대유구사방식、류서、자의조합이급수、미어의류（분대류층차화중류층차）5개특정적한어이자복합명사적구사규률진행료탐색，병취득료계단성성과。실험증명，재대규모한어어소수거고적기출상사용유전산법연구한어적구사규률시일충유효적방법。
The Chinese morpheme database is a large knowledge database aboutword formation. Chinese Natural Language Processing (NLP) needs to acquire word formation knowledge for Chinese compound words from this database. This paper describes the use of the genetic algorithm for Chinese NLP to study multi feature Chinese word formation. The analysis studied the combinatorial regulation of two character compound words with five features, the formation pattern, the sequence of morpheme category, the mean related to morphemes and the sense category of two morphemes (in major or medium categories). The results show that the genetic algorithm is an effective method for studying Chinese word formation based on the Chinese morpheme database.