Chinese Chunk Identification Based On Rule Extraction

Chunking can be used in the field of natural language processing which includes information extraction, Q&A System, and statements similarity comparison application systems and so on, it can also be applied to syntactic analysis. It plays an important role in these areas.Chinese functional chunk is identified mostly based on the rules which come from the syntactic analysis , it is mainly according to the role played by part-of-speech in parsing to mark the skeletal structure of the sentence. Compared to other methods, this method selects few features and has improved the recognition rate. Meanwhile, it also improves the recognition rate for the structure of multi-verb. After that the method of decision tree is used to identify Chinese functional chunk, this method improved the recognition rate on a certain extent. For Chinese basic chunk, The information based on the relationship between words, semantic relations and syntactic form are used to determine them. The method of the combination of rule extraction and statistical is used to determine Chinese basic chunk, this method effectively improved the recognition rate. This approach makes the Chinese basic chunk have a strong cohesion, it associates syntactic form and semantic content. The semantic disambiguation is also considered and has avoid them in a certain extent, rely the defects of solely on supplements.