2015 International Symposium of Computer Application and Information Technology (IsCait 2015)

277 / 1971-01-01 00:00:00

Address Information Segmentation Algorithm Based On Multiple Constraints On Bayesian Inference

geographic address,words segmentation,disambiguate,multiple constraints on Bayesian inference,segmentation accuracy

Traditional natural words segmentation technology is unable to efficiently tackle the issue of geographic address information. In order to solve this problem, multiple sources information from the Internet are used to constraint the input geographic address information. The multiple constraints model is introduced to explore the implied geographic information which can eliminate ambiguity of geographic information to enhance segmentation accuracy. The experiment results show that the proposed multiple constraints based Bayesian inference segmentation algorithm in this paper has better segmentation accuracy comparing with the latest Chinese academy of sciences segmentation algorithm, Institute of Computing Technology, Chinese Lexical Analysis System (ICTCLAS).