http://www.phpclasses.org/browse/package/2508.html

Word segmentation class using the maximum probability approach. Still in verifying now.

PS, I found that when using the same lexicon, the difference between the result segmented by maximum probability and reverse maximum match approach is less than 1%.

Simple Chinese Word Segmentation has been renamed to Fast Chinese Word Segmentation.

Whereafter, Fast Chinese Word Segmentation v0.05.2 was published.

今天看到 MeagerAvenger 发表的评论，让我看看 SmartTemplate。
其实几天以前我的模板就换成类似 Savant3 的了。自己写的 2KB 左右的类来驱动，觉得这样自定义方便，速度也快。