IDA
2001
Springer
14 years 4 months ago
2001
Springer
Abstract. We propose a new unsupervised training method for acquiring probability models that accurately segment Chinese character sequences into words. By constructing a core lexi...