Incrementally learning an Incremental parser
en-de
en-es
en-fr
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
'What is the relation between the language surface statistics and its hidden syntactic structure?' \\ To explore this, the author presents an unsupervised parser. Some of the properties of natural language that are used: Tree structures are skewed. Humans process language incrementally. Words have the Zipfian distribution. Bootstrapping in learning.