An Efficient Bayesian Framework Based Place Name Segmentation Algorithm for Geocoding System

In this paper, we propose an efficient Bayesian framework based place name segmentation algorithm which is suitable to be used in geocoding system. Firstly, architecture of the geocoding system is given, in which the effective address information can be output to users by the address quantization process based on quantization rules. To implement an effective geocoding system, the place name segmentation is one of the most important steps. The proposed place name segmentaton algorithm aims to calculate the logic by which the word segmentation problem can be solved effectively. We use the Bayesian framework to describe the statistical laws in the word segmentation problem, and the hypothesis space for the Bayesian framework consists of all the possible schemes of which probability are assigned to each word in the dictionary. Finally, word segmentation can be obtained by solving an optimization problem. Furthermore, experimental results show the effectiveness of the proposed algorithm.