Semi-supervised Multitask Learning for Sequence Labeling

ACL 2017 Marek Rei

We propose a sequence labeling framework with a secondary training objective, learning to predict surrounding words for every word in the dataset. This language modeling objective incentivises the system to learn general-purpose patterns of semantic and syntactic composition, which are also useful for improving accuracy on different sequence labeling tasks... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Grammatical Error Detection CoNLL-2014 A1 Bi-LSTM + LMcost (trained on FCE) F0.5 17.86 # 5
Grammatical Error Detection CoNLL-2014 A2 Bi-LSTM + LMcost (trained on FCE) F0.5 25.88 # 6
Grammatical Error Detection FCE Bi-LSTM + LMcost F0.5 48.48 # 3
Part-Of-Speech Tagging Penn Treebank Bi-LSTM + LMcost Accuracy 97.43 # 12