Shallow Semantic Parsing for Lexical Units in Chinese FrameNet
Xiaoyan Hao, Xiaoming Chang, Kaiying Liu · 2008
Recent years have seen a revived interest in semantic parsing by applying statistical and machine learning methods to semantically annotate corpora such as the FrameNet and the Proposition Bank. So far much of the research has been focused on English due to the lack of semantically annotated resources in other languages. This paper reports first results on semantic role labeling using a pre-release version of the Chinese FrameNet. In this work, we used the pre-release version of the Chinese FrameNet corpora to train a complete semantic role labeler. The labeler uses heuristic rules to look for lexical and semantic clues in the sentences. We evaluated the labeler by applying it to a small portion of the pre-release version of the Chinese FrameNet example corpus that was labeled manually. Overall, the labeler found the F-score of the correct Core roles 82.4% of the time, which is encouraging given the simplicity of its rules.