Developing a new statistical method for Chinese text segmentation

• Main Author: Dai, Yubin
• Other Authors: Khoo, Christopher Soo Guan
• Format: Theses and Dissertations
• Published: 2008
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Pattern recognition; DRNTU::Engineering::Computer science and engineering::Theory of computation::Analysis of algorithms and problem complexity
• Online Access: http://hdl.handle.net/10356/2614
• Institution: Nanyang Technological University

A new statistical formula for Chinese text segmentation called Contextual Information Formula (OF) was developed empirically for identifying 2 and 3-character words. It was developed by performing stepwise logistic regression using a sample of sentences that had been manually segmented. 300 sentences were used for model building and 100 sentences were set aside for model validation and evaluation. Relative frequencies, document frequencies, weighted document frequencies, within-document frequencies of characters, bigrams and trigrams were included in the study.