Chunking Tools Testing
1. Today, Tommy and I discussed the NLP problem with Wu. I tried chunking tools (C-OVA and YamCha) for getting the noun phrase and analyzed the tag to determine which tags should be merged. However, there is a very serious problem that the chunking tools contain the stemming process. So the result of noun phrase can’t achieve high precision. Such as, the chunking tools converts machine learning to machine learn.
2. Due to the chunking tools is an executive file. So I coded the simple program to run it on some testing plain text automatically for saving time.
1 comment:
I don't quite understand your problem here. I suppose you need to applly stemming after chuncking. If the chunking tool includes the stemming step, then you don't need additional stemming procedure. Why is this a problem?
Post a Comment