Abstract | ||
---|---|---|
Recent work on the design of automatic systems for semantic role labeling has shown that feature engineering is a com- plex task from a modeling and implemen- tation point of view. Tree kernels alleviate such complexity as kernel functions gener- ate features automatically and require less software development for data extraction. In this paper, we study several tree kernel approaches for both boundary detection and argument classification. The compar- ative experiments on Support Vector Ma- chines with such kernels on the CoNLL 2005 dataset show that very simple tree manipulations trigger automatic feature engineering that highly improves accuracy and efficiency in both phases. Moreover, the use of different classifiers for internal and pre-terminal nodes maintains the same accuracy and highly improves efficiency. |
Year | Venue | Keywords |
---|---|---|
2006 | Learning Structured Information@EACL | software development,semantic role labeling,support vector,kernel function |
Field | DocType | Citations |
Pattern recognition,Computer science,Support vector machine,Tree kernel,Feature engineering,Artificial intelligence,Kernel method,Machine learning,Software development,Semantic role labeling,Semantic computing,Kernel (statistics) | Conference | 12 |
PageRank | References | Authors |
0.62 | 15 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Alessandro Moschitti | 1 | 3262 | 177.68 |
daniele pighin | 2 | 289 | 18.72 |
Roberto Basili | 3 | 1308 | 155.68 |