Identification of Chinese organization name based on Chinese chunking

Ji-hao YIN,Xiao-zhong FAN,Pan-chao ZHAO,Jiang-de YU
DOI: https://doi.org/10.3969/j.issn.1006-7043.2006.z1.098
2006-01-01
Abstract:A simplified N-best cascade model is presented about Chinese organization names (ORG) identification. This model can do words segmentation, POS tagging, chunking and ORG identification. The N-best cascade method can not only control the errors propagation but also limit the search space. In the experiments, heuristic information and organization abbreviation processing are integrated into the model to achieve better experiment results. The precision and recall of Chinese ORG identification are 92.31% and 81.01% respectively in IEER99 test corpus.
What problem does this paper attempt to address?