基于概率统计模型的快递地址自动分类方法

快递货物在中转点向取送点分拣时需要人工判断收货地址所属取送点，为提高分拣的自动化程度和分拣速度，提出一种基于概率统计分类模型的快递地址自动分类方法。该方法以基于概率统计的地址分类模型为核心，通过统计出的最小地址要素与取送点的对应概率分布，对快递地址所属的取送点做出判断。在某快递公司提供的快递地址分类数据上的实验结果表明，该方法的自动分类准确率可达99%以上，每个地址的分类用时为0.43 ms。

Automatic Classification Approach of Express Address Based on Probability Statistical Model

In general, the delivery terminal that an express address belongs to is determined manually when sorting the goods at the express distribution center. In order to improve automation and speed, an automatic classification approach of express address based on the probability statistical model is proposed. The probability statistical model counts the probability distributions of the minimum address element, and determines the delivery terminal that the goods should be sent to. Experimental results based on the real data show that the classification accuracy of the approach reaches 99%, and classification speed is 0.43 ms per address.