Automatic TIMEX2 tagging of Chinese temporal information
暂无分享,去创建一个
To standardize Chinese temporal information tagging for automatic implementation, the Chinese draft of TIMEX2 tagging standard was modified. A Chinese TIMEX2 automatic tagging system(CTAT) based on regular-expressions was then developed. Three-level candidates are used to determine the reference time. Multiple-modules identify the temporal phrases from the input sentence. The results are sorted and collisions are eliminated to choose an optimum output. The F-measure of temporal expressions identification is 90.15% with a value normalization of 83.27%, similar to the levels of temporal information tagging systems in other languages.