A Simple and Efficient Model Pruning Method for Conditional Random Fields

Zhao, Hai; Kit, Chunyu

doi:10.1007/978-3-642-00831-3_14

Cited by 5 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If some features are pruned without significantly hurting the performance, these features may be less crucial or at least redundant with regard to the remaining features. Zhao & Kit (2009) have proposed a simple and efficient model pruning method for conditional random fields. A closer look at their experiments results on CRF based CWS helps us better understand the roles that different features play.…”

Section: The Minority Rulesmentioning

confidence: 99%

Distributional Evidence and Beyond: the Success and Limitations of Machine Learning in Chinese Word Segmentation

Gerdemann

2013

RCS

View full text Add to dashboard Cite

In this paper, we argue that the key to the success of the current state-of-art statistical learning algorithms for Chinese word segmentation (CWS) mostly lies in their optimal weighting of non-overlapping distributional evidence in the corpora. The utilization of distributional evidence is more essential than the learning algorithm. We further analyze the characteristics of distributional evidence for CWS, under the framework of Zipf's law and summarize the limitation of statistical learning in CWS as the feature absence problem, which may be apparent yet usually neglected. Making a connection between theoretical/empirical linguistics and CWS, we suggest that the study and development of a generative word formation system may be beneficial for both the science and engineering of CWS. We wrap up the discussion after reviewing some recent works that are already on this line.

show abstract

Section: The Minority Rulesmentioning

confidence: 99%

Distributional Evidence and Beyond: the Success and Limitations of Machine Learning in Chinese Word Segmentation

Gerdemann

2013

RCS

View full text Add to dashboard Cite

show abstract

“…The other is the conditional random fields (CRFs) model [23] for supervised segmentation via character tagging, conventionally trained only on a pre-segmented corpus. The latter is a state-of-the-art approach that has set new performance records in the field, as illustrated in [52,55], although its efficiency is yet to be further enhanced by various means [58,59]. All scores given by the goodness measures are discretized in the same way for use as feature values in the CRFs model.…”

Section: Introductionmentioning

confidence: 99%