Abstract: This paper presented a semi-supervised named entity recognition method based on conditional random field model for named entities in the military text,which aims at merging military named entities such as military appointment and military rank,military equipment,military supplies,military facilities,military institutions (including code designation) and military place names into a unified technical framework.The method establishes an efficient feature set according to the grammatical features of the military text,builds conditional random field model to identify the military named entities,and develops the method based on dictionary and the method based on rules to improve the results in turn.Experiments show that the method is able to complete the named entity recognition task in the military text well,and make F-value about testing the language material up to 90.9%,which is close to the level of named entity recognition in the commons area.

Key words: Military text,Named entity recognition,Conditional random field,Semi-supervised learning,Military information processing