欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出了一种基于词频的机械匹配自动分词算法,以长度为首优先,结合词频进行分词,未匹配字串进一步应用改进的正向和逆向的最大匹配法,结合熵率分词分别标注所有可能为词的元素。在匹配时完全忽略了5个字以上的词语,解决了随长度增大复杂度呈指数增长这一难题。实验表明,该方法改善了分词正确率,提高了切分效率。
英文摘要：
We present an automatic mechanical matching segmentation algorithm based on word frequency, in which the length is taken as the priority. The unmatched the strings are segmented with the improved forward-reverse maximum matching method. Combing with the entropy rate of word, the all possible elements are marked as words. The phase with over 5 words is ignored completely in the matching to reduce the complexity. The experimental results show that the word correct separating rate is improved and the segmentation efficiency is increased.
同期刊论文项目
同项目期刊论文
期刊信息
位置：