segmentation chinese word based on max boudary mix with CRF

The Sighan BakeOff result have release, and i receive the fifth place in Word Segmentation for Simplified Chinese open test (in compute) 、second place in Word Segmentation for Traditional Chinese open test (in compute、in medical、in finace、in literature). The result is satificate for me , but have some  pity that i don't receive the first place. i am earger for the first.

This year, there are 19 teams participate in this competition for the world, include CMU、Queensland University of Technology,SEG-IASL,pku, Fudan and  so on. Today Chinese word segmentation will focus on the cross-domain performance of Chinese word segmentation algorithms.

I think i was the only master to participate in this competition. and i was only one person. the result is get me many confidence.

I use the method "segmentation chinese word based on max boudary mix with CRF". the algorithm contain two step. first, i seg the Chinse word based on max bondary to generate the candition chunk, and to mark.The i use CRF to segment.

The detail please wait my paper pubilcing.

你可能感兴趣的:(word,Office,chinese,crf,segmentation,MIX)