Semantic Dependency Feature Selection
2008
This paper presents our solution for CoNLL 2008 shared task that jointly parses syntactic and semantic dependencies. The Maximum Entropy (ME) classifier has been selected to train the data used in this system. Also the Mutual Information (MI) model was utilized into feature selection of dependency labeling. Results show that the MI model allows the system to get better performance and less training hours.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
14
References
1
Citations
NaN
KQI