Abstract | ||
---|---|---|
This paper describes the submission of the National University of Singapore (NUS) to the HOO 2012 shared task. Our system uses a pipeline of confidence-weighted linear classifiers to correct determiner and preposition errors. Our system achieves the highest correction F1 score on the official test set among all 14 participating teams, based on gold-standard edits both before and after revision. |
Year | Venue | Keywords |
---|---|---|
2012 | BEA@NAACL-HLT | confidence-weighted linear classifier,modified treebank,error correction,highest correction,determiner error correction,shared task,preposition error,determiner error,system targets preposition,error correction system,recognition f-scores,f1 score,gold-standard edit,nara institute,national university,official test |
Field | DocType | Citations |
F1 score,Determiner,Computer science,Speech recognition,Natural language processing,Artificial intelligence,Test set | Conference | 5 |
PageRank | References | Authors |
0.44 | 24 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Daniel Dahlmeier | 1 | 460 | 29.67 |
Hwee Tou Ng | 2 | 4092 | 300.40 |
Eric Jun Feng Ng | 3 | 5 | 0.44 |