Using Similarity Measures to Select Pretraining Data for NER.

Xiang Dai Sarvnaz Karimi Ben Hachey Cécile Paris

Published in: NAACL-HLT (1) (2019)

Keyphrases

synthetic data
data sets
database
original data
similarity measure
data structure
raw data
data analysis
high quality
active learning
probability distribution
knowledge discovery
data processing
high dimensional data
training data
missing data
conditional random fields
experimental data
data distribution